diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7138 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.9815950920245395,
+  "eval_steps": 500,
+  "global_step": 1015,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0049079754601227,
+      "grad_norm": 2.422234535217285,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.5835,
+      "step": 1
+    },
+    {
+      "epoch": 0.0098159509202454,
+      "grad_norm": 0.8900374174118042,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.2599,
+      "step": 2
+    },
+    {
+      "epoch": 0.014723926380368098,
+      "grad_norm": 0.9694510102272034,
+      "learning_rate": 4.2e-06,
+      "loss": 0.1512,
+      "step": 3
+    },
+    {
+      "epoch": 0.0196319018404908,
+      "grad_norm": 2.383556604385376,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.5828,
+      "step": 4
+    },
+    {
+      "epoch": 0.024539877300613498,
+      "grad_norm": 0.5773197412490845,
+      "learning_rate": 7e-06,
+      "loss": 0.156,
+      "step": 5
+    },
+    {
+      "epoch": 0.029447852760736196,
+      "grad_norm": 0.8293938636779785,
+      "learning_rate": 6.993069306930693e-06,
+      "loss": 0.1723,
+      "step": 6
+    },
+    {
+      "epoch": 0.0343558282208589,
+      "grad_norm": 0.02590053714811802,
+      "learning_rate": 6.986138613861386e-06,
+      "loss": 0.0011,
+      "step": 7
+    },
+    {
+      "epoch": 0.0392638036809816,
+      "grad_norm": 0.627031683921814,
+      "learning_rate": 6.979207920792079e-06,
+      "loss": 0.2592,
+      "step": 8
+    },
+    {
+      "epoch": 0.044171779141104296,
+      "grad_norm": 6.028290271759033,
+      "learning_rate": 6.972277227722772e-06,
+      "loss": 0.7911,
+      "step": 9
+    },
+    {
+      "epoch": 0.049079754601226995,
+      "grad_norm": 1.6350908279418945,
+      "learning_rate": 6.965346534653465e-06,
+      "loss": 0.2866,
+      "step": 10
+    },
+    {
+      "epoch": 0.053987730061349694,
+      "grad_norm": 1.3962496519088745,
+      "learning_rate": 6.958415841584158e-06,
+      "loss": 0.2447,
+      "step": 11
+    },
+    {
+      "epoch": 0.05889570552147239,
+      "grad_norm": 0.385162889957428,
+      "learning_rate": 6.951485148514851e-06,
+      "loss": 0.1153,
+      "step": 12
+    },
+    {
+      "epoch": 0.0638036809815951,
+      "grad_norm": 0.6873381733894348,
+      "learning_rate": 6.9445544554455444e-06,
+      "loss": 0.6868,
+      "step": 13
+    },
+    {
+      "epoch": 0.0687116564417178,
+      "grad_norm": 0.3449646234512329,
+      "learning_rate": 6.9376237623762375e-06,
+      "loss": 0.1523,
+      "step": 14
+    },
+    {
+      "epoch": 0.0736196319018405,
+      "grad_norm": 2.022113084793091,
+      "learning_rate": 6.930693069306931e-06,
+      "loss": 0.3527,
+      "step": 15
+    },
+    {
+      "epoch": 0.0785276073619632,
+      "grad_norm": 0.661878764629364,
+      "learning_rate": 6.923762376237624e-06,
+      "loss": 0.2449,
+      "step": 16
+    },
+    {
+      "epoch": 0.0834355828220859,
+      "grad_norm": 0.7884008884429932,
+      "learning_rate": 6.916831683168317e-06,
+      "loss": 0.1496,
+      "step": 17
+    },
+    {
+      "epoch": 0.08834355828220859,
+      "grad_norm": 0.8798872828483582,
+      "learning_rate": 6.90990099009901e-06,
+      "loss": 0.2151,
+      "step": 18
+    },
+    {
+      "epoch": 0.09325153374233129,
+      "grad_norm": 1.3640304803848267,
+      "learning_rate": 6.902970297029703e-06,
+      "loss": 0.356,
+      "step": 19
+    },
+    {
+      "epoch": 0.09815950920245399,
+      "grad_norm": 1.1427268981933594,
+      "learning_rate": 6.896039603960396e-06,
+      "loss": 0.2645,
+      "step": 20
+    },
+    {
+      "epoch": 0.10306748466257669,
+      "grad_norm": 0.9606547951698303,
+      "learning_rate": 6.889108910891089e-06,
+      "loss": 0.3541,
+      "step": 21
+    },
+    {
+      "epoch": 0.10797546012269939,
+      "grad_norm": 0.23967993259429932,
+      "learning_rate": 6.882178217821782e-06,
+      "loss": 0.0407,
+      "step": 22
+    },
+    {
+      "epoch": 0.11288343558282209,
+      "grad_norm": 5.597607612609863,
+      "learning_rate": 6.875247524752475e-06,
+      "loss": 1.0892,
+      "step": 23
+    },
+    {
+      "epoch": 0.11779141104294479,
+      "grad_norm": 0.6763832569122314,
+      "learning_rate": 6.868316831683168e-06,
+      "loss": 0.4406,
+      "step": 24
+    },
+    {
+      "epoch": 0.12269938650306748,
+      "grad_norm": 5.346296310424805,
+      "learning_rate": 6.861386138613861e-06,
+      "loss": 1.2512,
+      "step": 25
+    },
+    {
+      "epoch": 0.1276073619631902,
+      "grad_norm": 0.9197913408279419,
+      "learning_rate": 6.854455445544554e-06,
+      "loss": 0.1855,
+      "step": 26
+    },
+    {
+      "epoch": 0.1325153374233129,
+      "grad_norm": 0.6145322918891907,
+      "learning_rate": 6.847524752475247e-06,
+      "loss": 0.0562,
+      "step": 27
+    },
+    {
+      "epoch": 0.1374233128834356,
+      "grad_norm": 1.5995681285858154,
+      "learning_rate": 6.8405940594059405e-06,
+      "loss": 0.405,
+      "step": 28
+    },
+    {
+      "epoch": 0.1423312883435583,
+      "grad_norm": 0.6084582805633545,
+      "learning_rate": 6.8336633663366335e-06,
+      "loss": 0.0953,
+      "step": 29
+    },
+    {
+      "epoch": 0.147239263803681,
+      "grad_norm": 0.09425808489322662,
+      "learning_rate": 6.826732673267327e-06,
+      "loss": 0.0128,
+      "step": 30
+    },
+    {
+      "epoch": 0.1521472392638037,
+      "grad_norm": 2.628293991088867,
+      "learning_rate": 6.81980198019802e-06,
+      "loss": 0.4032,
+      "step": 31
+    },
+    {
+      "epoch": 0.1570552147239264,
+      "grad_norm": 0.4179680347442627,
+      "learning_rate": 6.812871287128713e-06,
+      "loss": 0.0317,
+      "step": 32
+    },
+    {
+      "epoch": 0.1619631901840491,
+      "grad_norm": 1.106016755104065,
+      "learning_rate": 6.805940594059406e-06,
+      "loss": 0.1495,
+      "step": 33
+    },
+    {
+      "epoch": 0.1668711656441718,
+      "grad_norm": 0.3168316185474396,
+      "learning_rate": 6.799009900990099e-06,
+      "loss": 0.1907,
+      "step": 34
+    },
+    {
+      "epoch": 0.17177914110429449,
+      "grad_norm": 0.4211124777793884,
+      "learning_rate": 6.792079207920792e-06,
+      "loss": 0.0519,
+      "step": 35
+    },
+    {
+      "epoch": 0.17668711656441718,
+      "grad_norm": 1.7166955471038818,
+      "learning_rate": 6.785148514851485e-06,
+      "loss": 0.1693,
+      "step": 36
+    },
+    {
+      "epoch": 0.18159509202453988,
+      "grad_norm": 1.6126145124435425,
+      "learning_rate": 6.778217821782178e-06,
+      "loss": 0.5255,
+      "step": 37
+    },
+    {
+      "epoch": 0.18650306748466258,
+      "grad_norm": 0.6241395473480225,
+      "learning_rate": 6.771287128712871e-06,
+      "loss": 0.2787,
+      "step": 38
+    },
+    {
+      "epoch": 0.19141104294478528,
+      "grad_norm": 0.8319867253303528,
+      "learning_rate": 6.764356435643564e-06,
+      "loss": 0.0874,
+      "step": 39
+    },
+    {
+      "epoch": 0.19631901840490798,
+      "grad_norm": 2.6518406867980957,
+      "learning_rate": 6.757425742574257e-06,
+      "loss": 0.1322,
+      "step": 40
+    },
+    {
+      "epoch": 0.20122699386503068,
+      "grad_norm": 0.830451488494873,
+      "learning_rate": 6.75049504950495e-06,
+      "loss": 0.0785,
+      "step": 41
+    },
+    {
+      "epoch": 0.20613496932515338,
+      "grad_norm": 0.9671738147735596,
+      "learning_rate": 6.7435643564356434e-06,
+      "loss": 0.1281,
+      "step": 42
+    },
+    {
+      "epoch": 0.21104294478527608,
+      "grad_norm": 0.2914385497570038,
+      "learning_rate": 6.7366336633663365e-06,
+      "loss": 0.0565,
+      "step": 43
+    },
+    {
+      "epoch": 0.21595092024539878,
+      "grad_norm": 1.4892910718917847,
+      "learning_rate": 6.7297029702970296e-06,
+      "loss": 0.3326,
+      "step": 44
+    },
+    {
+      "epoch": 0.22085889570552147,
+      "grad_norm": 0.3593141734600067,
+      "learning_rate": 6.722772277227723e-06,
+      "loss": 0.0499,
+      "step": 45
+    },
+    {
+      "epoch": 0.22576687116564417,
+      "grad_norm": 0.8170046210289001,
+      "learning_rate": 6.715841584158416e-06,
+      "loss": 0.028,
+      "step": 46
+    },
+    {
+      "epoch": 0.23067484662576687,
+      "grad_norm": 1.7005876302719116,
+      "learning_rate": 6.708910891089109e-06,
+      "loss": 0.1811,
+      "step": 47
+    },
+    {
+      "epoch": 0.23558282208588957,
+      "grad_norm": 3.5889787673950195,
+      "learning_rate": 6.701980198019802e-06,
+      "loss": 0.5892,
+      "step": 48
+    },
+    {
+      "epoch": 0.24049079754601227,
+      "grad_norm": 0.26327264308929443,
+      "learning_rate": 6.695049504950495e-06,
+      "loss": 0.0134,
+      "step": 49
+    },
+    {
+      "epoch": 0.24539877300613497,
+      "grad_norm": 3.6891543865203857,
+      "learning_rate": 6.688118811881188e-06,
+      "loss": 0.2651,
+      "step": 50
+    },
+    {
+      "epoch": 0.25030674846625767,
+      "grad_norm": 0.229234978556633,
+      "learning_rate": 6.681188118811881e-06,
+      "loss": 0.0307,
+      "step": 51
+    },
+    {
+      "epoch": 0.2552147239263804,
+      "grad_norm": 0.4476153552532196,
+      "learning_rate": 6.674257425742574e-06,
+      "loss": 0.0974,
+      "step": 52
+    },
+    {
+      "epoch": 0.26012269938650306,
+      "grad_norm": 0.8651221394538879,
+      "learning_rate": 6.667326732673267e-06,
+      "loss": 0.136,
+      "step": 53
+    },
+    {
+      "epoch": 0.2650306748466258,
+      "grad_norm": 0.8453481197357178,
+      "learning_rate": 6.66039603960396e-06,
+      "loss": 0.0714,
+      "step": 54
+    },
+    {
+      "epoch": 0.26993865030674846,
+      "grad_norm": 0.8162591457366943,
+      "learning_rate": 6.653465346534653e-06,
+      "loss": 0.0607,
+      "step": 55
+    },
+    {
+      "epoch": 0.2748466257668712,
+      "grad_norm": 1.0115143060684204,
+      "learning_rate": 6.646534653465346e-06,
+      "loss": 0.0377,
+      "step": 56
+    },
+    {
+      "epoch": 0.27975460122699386,
+      "grad_norm": 0.18834809958934784,
+      "learning_rate": 6.6396039603960395e-06,
+      "loss": 0.0062,
+      "step": 57
+    },
+    {
+      "epoch": 0.2846625766871166,
+      "grad_norm": 0.06441894173622131,
+      "learning_rate": 6.6326732673267325e-06,
+      "loss": 0.0094,
+      "step": 58
+    },
+    {
+      "epoch": 0.28957055214723926,
+      "grad_norm": 1.1535841226577759,
+      "learning_rate": 6.625742574257426e-06,
+      "loss": 0.0234,
+      "step": 59
+    },
+    {
+      "epoch": 0.294478527607362,
+      "grad_norm": 0.1546783298254013,
+      "learning_rate": 6.618811881188119e-06,
+      "loss": 0.0192,
+      "step": 60
+    },
+    {
+      "epoch": 0.29938650306748466,
+      "grad_norm": 6.198948383331299,
+      "learning_rate": 6.611881188118812e-06,
+      "loss": 0.0637,
+      "step": 61
+    },
+    {
+      "epoch": 0.3042944785276074,
+      "grad_norm": 1.3115315437316895,
+      "learning_rate": 6.604950495049505e-06,
+      "loss": 0.074,
+      "step": 62
+    },
+    {
+      "epoch": 0.30920245398773005,
+      "grad_norm": 0.5492228269577026,
+      "learning_rate": 6.598019801980198e-06,
+      "loss": 0.0291,
+      "step": 63
+    },
+    {
+      "epoch": 0.3141104294478528,
+      "grad_norm": 0.46307000517845154,
+      "learning_rate": 6.591089108910891e-06,
+      "loss": 0.1191,
+      "step": 64
+    },
+    {
+      "epoch": 0.31901840490797545,
+      "grad_norm": 0.16621133685112,
+      "learning_rate": 6.584158415841584e-06,
+      "loss": 0.0317,
+      "step": 65
+    },
+    {
+      "epoch": 0.3239263803680982,
+      "grad_norm": 0.56168532371521,
+      "learning_rate": 6.577227722772277e-06,
+      "loss": 0.0797,
+      "step": 66
+    },
+    {
+      "epoch": 0.32883435582822085,
+      "grad_norm": 0.21208958327770233,
+      "learning_rate": 6.57029702970297e-06,
+      "loss": 0.0039,
+      "step": 67
+    },
+    {
+      "epoch": 0.3337423312883436,
+      "grad_norm": 0.5419512391090393,
+      "learning_rate": 6.563366336633663e-06,
+      "loss": 0.0269,
+      "step": 68
+    },
+    {
+      "epoch": 0.33865030674846625,
+      "grad_norm": 0.278899222612381,
+      "learning_rate": 6.556435643564357e-06,
+      "loss": 0.0593,
+      "step": 69
+    },
+    {
+      "epoch": 0.34355828220858897,
+      "grad_norm": 0.6312870979309082,
+      "learning_rate": 6.54950495049505e-06,
+      "loss": 0.0751,
+      "step": 70
+    },
+    {
+      "epoch": 0.34846625766871164,
+      "grad_norm": 0.19221121072769165,
+      "learning_rate": 6.542574257425743e-06,
+      "loss": 0.0037,
+      "step": 71
+    },
+    {
+      "epoch": 0.35337423312883437,
+      "grad_norm": 0.29226556420326233,
+      "learning_rate": 6.5356435643564355e-06,
+      "loss": 0.0417,
+      "step": 72
+    },
+    {
+      "epoch": 0.35828220858895704,
+      "grad_norm": 0.32889777421951294,
+      "learning_rate": 6.5287128712871286e-06,
+      "loss": 0.0767,
+      "step": 73
+    },
+    {
+      "epoch": 0.36319018404907977,
+      "grad_norm": 0.665745735168457,
+      "learning_rate": 6.521782178217822e-06,
+      "loss": 0.1935,
+      "step": 74
+    },
+    {
+      "epoch": 0.36809815950920244,
+      "grad_norm": 0.10821861028671265,
+      "learning_rate": 6.514851485148515e-06,
+      "loss": 0.0087,
+      "step": 75
+    },
+    {
+      "epoch": 0.37300613496932516,
+      "grad_norm": 1.0068309307098389,
+      "learning_rate": 6.507920792079208e-06,
+      "loss": 0.0232,
+      "step": 76
+    },
+    {
+      "epoch": 0.37791411042944784,
+      "grad_norm": 0.03962863236665726,
+      "learning_rate": 6.500990099009901e-06,
+      "loss": 0.0051,
+      "step": 77
+    },
+    {
+      "epoch": 0.38282208588957056,
+      "grad_norm": 9.40488052368164,
+      "learning_rate": 6.494059405940594e-06,
+      "loss": 0.2292,
+      "step": 78
+    },
+    {
+      "epoch": 0.38773006134969323,
+      "grad_norm": 0.49305009841918945,
+      "learning_rate": 6.487128712871287e-06,
+      "loss": 0.0924,
+      "step": 79
+    },
+    {
+      "epoch": 0.39263803680981596,
+      "grad_norm": 0.72562575340271,
+      "learning_rate": 6.48019801980198e-06,
+      "loss": 0.0549,
+      "step": 80
+    },
+    {
+      "epoch": 0.39754601226993863,
+      "grad_norm": 0.015473410487174988,
+      "learning_rate": 6.473267326732673e-06,
+      "loss": 0.0017,
+      "step": 81
+    },
+    {
+      "epoch": 0.40245398773006136,
+      "grad_norm": 0.13165496289730072,
+      "learning_rate": 6.466336633663366e-06,
+      "loss": 0.0068,
+      "step": 82
+    },
+    {
+      "epoch": 0.40736196319018403,
+      "grad_norm": 0.42630520462989807,
+      "learning_rate": 6.459405940594059e-06,
+      "loss": 0.0927,
+      "step": 83
+    },
+    {
+      "epoch": 0.41226993865030676,
+      "grad_norm": 0.059238456189632416,
+      "learning_rate": 6.452475247524752e-06,
+      "loss": 0.0036,
+      "step": 84
+    },
+    {
+      "epoch": 0.4171779141104294,
+      "grad_norm": 0.8484250903129578,
+      "learning_rate": 6.445544554455445e-06,
+      "loss": 0.2075,
+      "step": 85
+    },
+    {
+      "epoch": 0.42208588957055215,
+      "grad_norm": 0.3800269067287445,
+      "learning_rate": 6.4386138613861384e-06,
+      "loss": 0.1287,
+      "step": 86
+    },
+    {
+      "epoch": 0.4269938650306748,
+      "grad_norm": 0.002063202438876033,
+      "learning_rate": 6.4316831683168315e-06,
+      "loss": 0.0008,
+      "step": 87
+    },
+    {
+      "epoch": 0.43190184049079755,
+      "grad_norm": 0.05038560926914215,
+      "learning_rate": 6.424752475247525e-06,
+      "loss": 0.0017,
+      "step": 88
+    },
+    {
+      "epoch": 0.4368098159509202,
+      "grad_norm": 0.0683569684624672,
+      "learning_rate": 6.417821782178218e-06,
+      "loss": 0.0035,
+      "step": 89
+    },
+    {
+      "epoch": 0.44171779141104295,
+      "grad_norm": 0.028999239206314087,
+      "learning_rate": 6.410891089108911e-06,
+      "loss": 0.002,
+      "step": 90
+    },
+    {
+      "epoch": 0.4466257668711656,
+      "grad_norm": 0.1129206195473671,
+      "learning_rate": 6.403960396039604e-06,
+      "loss": 0.0045,
+      "step": 91
+    },
+    {
+      "epoch": 0.45153374233128835,
+      "grad_norm": 0.012081787921488285,
+      "learning_rate": 6.397029702970297e-06,
+      "loss": 0.0016,
+      "step": 92
+    },
+    {
+      "epoch": 0.456441717791411,
+      "grad_norm": 0.29454007744789124,
+      "learning_rate": 6.39009900990099e-06,
+      "loss": 0.0047,
+      "step": 93
+    },
+    {
+      "epoch": 0.46134969325153374,
+      "grad_norm": 0.342790812253952,
+      "learning_rate": 6.383168316831683e-06,
+      "loss": 0.0182,
+      "step": 94
+    },
+    {
+      "epoch": 0.4662576687116564,
+      "grad_norm": 0.3073101341724396,
+      "learning_rate": 6.376237623762376e-06,
+      "loss": 0.1144,
+      "step": 95
+    },
+    {
+      "epoch": 0.47116564417177914,
+      "grad_norm": 0.028703156858682632,
+      "learning_rate": 6.369306930693069e-06,
+      "loss": 0.0017,
+      "step": 96
+    },
+    {
+      "epoch": 0.47607361963190187,
+      "grad_norm": 0.2344651073217392,
+      "learning_rate": 6.362376237623762e-06,
+      "loss": 0.1945,
+      "step": 97
+    },
+    {
+      "epoch": 0.48098159509202454,
+      "grad_norm": 0.4256807565689087,
+      "learning_rate": 6.355445544554455e-06,
+      "loss": 0.0432,
+      "step": 98
+    },
+    {
+      "epoch": 0.48588957055214727,
+      "grad_norm": 0.690371572971344,
+      "learning_rate": 6.348514851485148e-06,
+      "loss": 0.1223,
+      "step": 99
+    },
+    {
+      "epoch": 0.49079754601226994,
+      "grad_norm": 0.006785046309232712,
+      "learning_rate": 6.341584158415841e-06,
+      "loss": 0.0009,
+      "step": 100
+    },
+    {
+      "epoch": 0.49570552147239266,
+      "grad_norm": 0.3789716362953186,
+      "learning_rate": 6.3346534653465345e-06,
+      "loss": 0.0753,
+      "step": 101
+    },
+    {
+      "epoch": 0.5006134969325153,
+      "grad_norm": 0.2657737731933594,
+      "learning_rate": 6.3277227722772275e-06,
+      "loss": 0.0068,
+      "step": 102
+    },
+    {
+      "epoch": 0.505521472392638,
+      "grad_norm": 0.0025652130134403706,
+      "learning_rate": 6.320792079207921e-06,
+      "loss": 0.0005,
+      "step": 103
+    },
+    {
+      "epoch": 0.5104294478527608,
+      "grad_norm": 0.026512114331126213,
+      "learning_rate": 6.313861386138614e-06,
+      "loss": 0.0012,
+      "step": 104
+    },
+    {
+      "epoch": 0.5153374233128835,
+      "grad_norm": 0.020907960832118988,
+      "learning_rate": 6.306930693069307e-06,
+      "loss": 0.002,
+      "step": 105
+    },
+    {
+      "epoch": 0.5202453987730061,
+      "grad_norm": 0.024951621890068054,
+      "learning_rate": 6.3e-06,
+      "loss": 0.0019,
+      "step": 106
+    },
+    {
+      "epoch": 0.5251533742331288,
+      "grad_norm": 0.03137391433119774,
+      "learning_rate": 6.293069306930693e-06,
+      "loss": 0.0022,
+      "step": 107
+    },
+    {
+      "epoch": 0.5300613496932516,
+      "grad_norm": 0.01158374547958374,
+      "learning_rate": 6.286138613861386e-06,
+      "loss": 0.002,
+      "step": 108
+    },
+    {
+      "epoch": 0.5349693251533743,
+      "grad_norm": 0.20386900007724762,
+      "learning_rate": 6.279207920792079e-06,
+      "loss": 0.0438,
+      "step": 109
+    },
+    {
+      "epoch": 0.5398773006134969,
+      "grad_norm": 0.21870765089988708,
+      "learning_rate": 6.272277227722772e-06,
+      "loss": 0.0147,
+      "step": 110
+    },
+    {
+      "epoch": 0.5447852760736196,
+      "grad_norm": 0.01832536980509758,
+      "learning_rate": 6.265346534653465e-06,
+      "loss": 0.0013,
+      "step": 111
+    },
+    {
+      "epoch": 0.5496932515337424,
+      "grad_norm": 1.1292037963867188,
+      "learning_rate": 6.258415841584158e-06,
+      "loss": 0.0233,
+      "step": 112
+    },
+    {
+      "epoch": 0.554601226993865,
+      "grad_norm": 0.23187655210494995,
+      "learning_rate": 6.251485148514851e-06,
+      "loss": 0.0964,
+      "step": 113
+    },
+    {
+      "epoch": 0.5595092024539877,
+      "grad_norm": 0.05976148694753647,
+      "learning_rate": 6.244554455445544e-06,
+      "loss": 0.0026,
+      "step": 114
+    },
+    {
+      "epoch": 0.5644171779141104,
+      "grad_norm": 0.2549298107624054,
+      "learning_rate": 6.2376237623762374e-06,
+      "loss": 0.0204,
+      "step": 115
+    },
+    {
+      "epoch": 0.5693251533742332,
+      "grad_norm": 0.5240129232406616,
+      "learning_rate": 6.2306930693069305e-06,
+      "loss": 0.0091,
+      "step": 116
+    },
+    {
+      "epoch": 0.5742331288343558,
+      "grad_norm": 0.008397325873374939,
+      "learning_rate": 6.2237623762376236e-06,
+      "loss": 0.0031,
+      "step": 117
+    },
+    {
+      "epoch": 0.5791411042944785,
+      "grad_norm": 0.17320525646209717,
+      "learning_rate": 6.216831683168317e-06,
+      "loss": 0.0696,
+      "step": 118
+    },
+    {
+      "epoch": 0.5840490797546012,
+      "grad_norm": 0.8707240223884583,
+      "learning_rate": 6.20990099009901e-06,
+      "loss": 0.0563,
+      "step": 119
+    },
+    {
+      "epoch": 0.588957055214724,
+      "grad_norm": 0.03828573226928711,
+      "learning_rate": 6.202970297029703e-06,
+      "loss": 0.0025,
+      "step": 120
+    },
+    {
+      "epoch": 0.5938650306748466,
+      "grad_norm": 0.39433979988098145,
+      "learning_rate": 6.196039603960396e-06,
+      "loss": 0.0286,
+      "step": 121
+    },
+    {
+      "epoch": 0.5987730061349693,
+      "grad_norm": 0.24125856161117554,
+      "learning_rate": 6.189108910891089e-06,
+      "loss": 0.0474,
+      "step": 122
+    },
+    {
+      "epoch": 0.603680981595092,
+      "grad_norm": 0.3200415074825287,
+      "learning_rate": 6.182178217821782e-06,
+      "loss": 0.0138,
+      "step": 123
+    },
+    {
+      "epoch": 0.6085889570552148,
+      "grad_norm": 0.016078324988484383,
+      "learning_rate": 6.175247524752475e-06,
+      "loss": 0.0015,
+      "step": 124
+    },
+    {
+      "epoch": 0.6134969325153374,
+      "grad_norm": 0.023426106199622154,
+      "learning_rate": 6.168316831683168e-06,
+      "loss": 0.0025,
+      "step": 125
+    },
+    {
+      "epoch": 0.6184049079754601,
+      "grad_norm": 0.255380779504776,
+      "learning_rate": 6.161386138613861e-06,
+      "loss": 0.0091,
+      "step": 126
+    },
+    {
+      "epoch": 0.6233128834355828,
+      "grad_norm": 0.02590188756585121,
+      "learning_rate": 6.154455445544554e-06,
+      "loss": 0.0025,
+      "step": 127
+    },
+    {
+      "epoch": 0.6282208588957056,
+      "grad_norm": 1.4449337720870972,
+      "learning_rate": 6.147524752475247e-06,
+      "loss": 0.1513,
+      "step": 128
+    },
+    {
+      "epoch": 0.6331288343558282,
+      "grad_norm": 0.27828818559646606,
+      "learning_rate": 6.14059405940594e-06,
+      "loss": 0.0542,
+      "step": 129
+    },
+    {
+      "epoch": 0.6380368098159509,
+      "grad_norm": 0.0807948037981987,
+      "learning_rate": 6.1336633663366335e-06,
+      "loss": 0.0033,
+      "step": 130
+    },
+    {
+      "epoch": 0.6429447852760736,
+      "grad_norm": 0.5472558736801147,
+      "learning_rate": 6.1267326732673265e-06,
+      "loss": 0.1161,
+      "step": 131
+    },
+    {
+      "epoch": 0.6478527607361964,
+      "grad_norm": 0.18089702725410461,
+      "learning_rate": 6.11980198019802e-06,
+      "loss": 0.0464,
+      "step": 132
+    },
+    {
+      "epoch": 0.652760736196319,
+      "grad_norm": 0.025230491533875465,
+      "learning_rate": 6.112871287128713e-06,
+      "loss": 0.0016,
+      "step": 133
+    },
+    {
+      "epoch": 0.6576687116564417,
+      "grad_norm": 0.2929389178752899,
+      "learning_rate": 6.105940594059406e-06,
+      "loss": 0.0404,
+      "step": 134
+    },
+    {
+      "epoch": 0.6625766871165644,
+      "grad_norm": 0.2341107577085495,
+      "learning_rate": 6.099009900990099e-06,
+      "loss": 0.0401,
+      "step": 135
+    },
+    {
+      "epoch": 0.6674846625766871,
+      "grad_norm": 0.11038243025541306,
+      "learning_rate": 6.092079207920792e-06,
+      "loss": 0.0045,
+      "step": 136
+    },
+    {
+      "epoch": 0.6723926380368098,
+      "grad_norm": 0.19471372663974762,
+      "learning_rate": 6.085148514851485e-06,
+      "loss": 0.0296,
+      "step": 137
+    },
+    {
+      "epoch": 0.6773006134969325,
+      "grad_norm": 0.3493005037307739,
+      "learning_rate": 6.078217821782178e-06,
+      "loss": 0.0435,
+      "step": 138
+    },
+    {
+      "epoch": 0.6822085889570552,
+      "grad_norm": 0.2767382860183716,
+      "learning_rate": 6.071287128712871e-06,
+      "loss": 0.0392,
+      "step": 139
+    },
+    {
+      "epoch": 0.6871165644171779,
+      "grad_norm": 0.2326585054397583,
+      "learning_rate": 6.064356435643564e-06,
+      "loss": 0.0372,
+      "step": 140
+    },
+    {
+      "epoch": 0.6920245398773006,
+      "grad_norm": 0.023760871961712837,
+      "learning_rate": 6.057425742574257e-06,
+      "loss": 0.0017,
+      "step": 141
+    },
+    {
+      "epoch": 0.6969325153374233,
+      "grad_norm": 0.02614918164908886,
+      "learning_rate": 6.05049504950495e-06,
+      "loss": 0.0019,
+      "step": 142
+    },
+    {
+      "epoch": 0.701840490797546,
+      "grad_norm": 0.21519997715950012,
+      "learning_rate": 6.043564356435643e-06,
+      "loss": 0.0249,
+      "step": 143
+    },
+    {
+      "epoch": 0.7067484662576687,
+      "grad_norm": 0.04296768456697464,
+      "learning_rate": 6.0366336633663364e-06,
+      "loss": 0.002,
+      "step": 144
+    },
+    {
+      "epoch": 0.7116564417177914,
+      "grad_norm": 0.0280557032674551,
+      "learning_rate": 6.0297029702970295e-06,
+      "loss": 0.0012,
+      "step": 145
+    },
+    {
+      "epoch": 0.7165644171779141,
+      "grad_norm": 0.016091475263237953,
+      "learning_rate": 6.0227722772277226e-06,
+      "loss": 0.001,
+      "step": 146
+    },
+    {
+      "epoch": 0.7214723926380369,
+      "grad_norm": 0.4104843735694885,
+      "learning_rate": 6.015841584158416e-06,
+      "loss": 0.0495,
+      "step": 147
+    },
+    {
+      "epoch": 0.7263803680981595,
+      "grad_norm": 0.28999537229537964,
+      "learning_rate": 6.008910891089109e-06,
+      "loss": 0.0377,
+      "step": 148
+    },
+    {
+      "epoch": 0.7312883435582822,
+      "grad_norm": 0.005074529442936182,
+      "learning_rate": 6.001980198019802e-06,
+      "loss": 0.0006,
+      "step": 149
+    },
+    {
+      "epoch": 0.7361963190184049,
+      "grad_norm": 0.031069966033101082,
+      "learning_rate": 5.995049504950495e-06,
+      "loss": 0.0013,
+      "step": 150
+    },
+    {
+      "epoch": 0.7411042944785277,
+      "grad_norm": 0.0037307555321604013,
+      "learning_rate": 5.988118811881188e-06,
+      "loss": 0.0004,
+      "step": 151
+    },
+    {
+      "epoch": 0.7460122699386503,
+      "grad_norm": 0.05543583631515503,
+      "learning_rate": 5.981188118811881e-06,
+      "loss": 0.0056,
+      "step": 152
+    },
+    {
+      "epoch": 0.750920245398773,
+      "grad_norm": 0.15376004576683044,
+      "learning_rate": 5.974257425742574e-06,
+      "loss": 0.0072,
+      "step": 153
+    },
+    {
+      "epoch": 0.7558282208588957,
+      "grad_norm": 0.17796596884727478,
+      "learning_rate": 5.967326732673267e-06,
+      "loss": 0.0329,
+      "step": 154
+    },
+    {
+      "epoch": 0.7607361963190185,
+      "grad_norm": 0.15298157930374146,
+      "learning_rate": 5.96039603960396e-06,
+      "loss": 0.014,
+      "step": 155
+    },
+    {
+      "epoch": 0.7656441717791411,
+      "grad_norm": 0.017613040283322334,
+      "learning_rate": 5.953465346534653e-06,
+      "loss": 0.0016,
+      "step": 156
+    },
+    {
+      "epoch": 0.7705521472392638,
+      "grad_norm": 0.33242282271385193,
+      "learning_rate": 5.946534653465346e-06,
+      "loss": 0.0333,
+      "step": 157
+    },
+    {
+      "epoch": 0.7754601226993865,
+      "grad_norm": 0.010239041410386562,
+      "learning_rate": 5.939603960396039e-06,
+      "loss": 0.0005,
+      "step": 158
+    },
+    {
+      "epoch": 0.7803680981595092,
+      "grad_norm": 0.01592794805765152,
+      "learning_rate": 5.9326732673267325e-06,
+      "loss": 0.0025,
+      "step": 159
+    },
+    {
+      "epoch": 0.7852760736196319,
+      "grad_norm": 0.3407064974308014,
+      "learning_rate": 5.9257425742574255e-06,
+      "loss": 0.0106,
+      "step": 160
+    },
+    {
+      "epoch": 0.7901840490797546,
+      "grad_norm": 0.11607719957828522,
+      "learning_rate": 5.918811881188119e-06,
+      "loss": 0.0197,
+      "step": 161
+    },
+    {
+      "epoch": 0.7950920245398773,
+      "grad_norm": 0.2373722642660141,
+      "learning_rate": 5.911881188118812e-06,
+      "loss": 0.009,
+      "step": 162
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.013574715703725815,
+      "learning_rate": 5.904950495049505e-06,
+      "loss": 0.0012,
+      "step": 163
+    },
+    {
+      "epoch": 0.8049079754601227,
+      "grad_norm": 0.1712418496608734,
+      "learning_rate": 5.898019801980198e-06,
+      "loss": 0.0033,
+      "step": 164
+    },
+    {
+      "epoch": 0.8098159509202454,
+      "grad_norm": 0.06470532715320587,
+      "learning_rate": 5.891089108910891e-06,
+      "loss": 0.0022,
+      "step": 165
+    },
+    {
+      "epoch": 0.8147239263803681,
+      "grad_norm": 0.2027478665113449,
+      "learning_rate": 5.884158415841584e-06,
+      "loss": 0.0075,
+      "step": 166
+    },
+    {
+      "epoch": 0.8196319018404908,
+      "grad_norm": 0.018224777653813362,
+      "learning_rate": 5.877227722772277e-06,
+      "loss": 0.0022,
+      "step": 167
+    },
+    {
+      "epoch": 0.8245398773006135,
+      "grad_norm": 0.2861013114452362,
+      "learning_rate": 5.87029702970297e-06,
+      "loss": 0.0213,
+      "step": 168
+    },
+    {
+      "epoch": 0.8294478527607362,
+      "grad_norm": 0.1993497610092163,
+      "learning_rate": 5.863366336633663e-06,
+      "loss": 0.0047,
+      "step": 169
+    },
+    {
+      "epoch": 0.8343558282208589,
+      "grad_norm": 0.010481802746653557,
+      "learning_rate": 5.856435643564356e-06,
+      "loss": 0.001,
+      "step": 170
+    },
+    {
+      "epoch": 0.8392638036809816,
+      "grad_norm": 0.018094172701239586,
+      "learning_rate": 5.849504950495049e-06,
+      "loss": 0.0018,
+      "step": 171
+    },
+    {
+      "epoch": 0.8441717791411043,
+      "grad_norm": 0.06920409202575684,
+      "learning_rate": 5.842574257425742e-06,
+      "loss": 0.0042,
+      "step": 172
+    },
+    {
+      "epoch": 0.849079754601227,
+      "grad_norm": 0.18376444280147552,
+      "learning_rate": 5.835643564356435e-06,
+      "loss": 0.0066,
+      "step": 173
+    },
+    {
+      "epoch": 0.8539877300613496,
+      "grad_norm": 0.1790069192647934,
+      "learning_rate": 5.8287128712871285e-06,
+      "loss": 0.0147,
+      "step": 174
+    },
+    {
+      "epoch": 0.8588957055214724,
+      "grad_norm": 0.018982654437422752,
+      "learning_rate": 5.8217821782178216e-06,
+      "loss": 0.003,
+      "step": 175
+    },
+    {
+      "epoch": 0.8638036809815951,
+      "grad_norm": 0.13061672449111938,
+      "learning_rate": 5.814851485148515e-06,
+      "loss": 0.0016,
+      "step": 176
+    },
+    {
+      "epoch": 0.8687116564417178,
+      "grad_norm": 0.0860314592719078,
+      "learning_rate": 5.807920792079208e-06,
+      "loss": 0.0047,
+      "step": 177
+    },
+    {
+      "epoch": 0.8736196319018404,
+      "grad_norm": 0.05230861157178879,
+      "learning_rate": 5.800990099009901e-06,
+      "loss": 0.0076,
+      "step": 178
+    },
+    {
+      "epoch": 0.8785276073619632,
+      "grad_norm": 0.3662849962711334,
+      "learning_rate": 5.794059405940594e-06,
+      "loss": 0.026,
+      "step": 179
+    },
+    {
+      "epoch": 0.8834355828220859,
+      "grad_norm": 0.2983264923095703,
+      "learning_rate": 5.787128712871287e-06,
+      "loss": 0.0203,
+      "step": 180
+    },
+    {
+      "epoch": 0.8883435582822086,
+      "grad_norm": 0.0335027277469635,
+      "learning_rate": 5.78019801980198e-06,
+      "loss": 0.0022,
+      "step": 181
+    },
+    {
+      "epoch": 0.8932515337423312,
+      "grad_norm": 0.4801461100578308,
+      "learning_rate": 5.773267326732673e-06,
+      "loss": 0.0082,
+      "step": 182
+    },
+    {
+      "epoch": 0.898159509202454,
+      "grad_norm": 0.014546169899404049,
+      "learning_rate": 5.766336633663366e-06,
+      "loss": 0.0031,
+      "step": 183
+    },
+    {
+      "epoch": 0.9030674846625767,
+      "grad_norm": 0.05124737694859505,
+      "learning_rate": 5.759405940594059e-06,
+      "loss": 0.002,
+      "step": 184
+    },
+    {
+      "epoch": 0.9079754601226994,
+      "grad_norm": 0.01893136277794838,
+      "learning_rate": 5.752475247524752e-06,
+      "loss": 0.0013,
+      "step": 185
+    },
+    {
+      "epoch": 0.912883435582822,
+      "grad_norm": 0.15197233855724335,
+      "learning_rate": 5.745544554455445e-06,
+      "loss": 0.0026,
+      "step": 186
+    },
+    {
+      "epoch": 0.9177914110429448,
+      "grad_norm": 0.8003377318382263,
+      "learning_rate": 5.738613861386138e-06,
+      "loss": 0.0146,
+      "step": 187
+    },
+    {
+      "epoch": 0.9226993865030675,
+      "grad_norm": 0.023129871115088463,
+      "learning_rate": 5.7316831683168314e-06,
+      "loss": 0.0035,
+      "step": 188
+    },
+    {
+      "epoch": 0.9276073619631902,
+      "grad_norm": 0.5904732346534729,
+      "learning_rate": 5.7247524752475245e-06,
+      "loss": 0.0223,
+      "step": 189
+    },
+    {
+      "epoch": 0.9325153374233128,
+      "grad_norm": 0.02393723465502262,
+      "learning_rate": 5.717821782178218e-06,
+      "loss": 0.0017,
+      "step": 190
+    },
+    {
+      "epoch": 0.9374233128834356,
+      "grad_norm": 0.2537831664085388,
+      "learning_rate": 5.710891089108911e-06,
+      "loss": 0.0113,
+      "step": 191
+    },
+    {
+      "epoch": 0.9423312883435583,
+      "grad_norm": 0.005700583104044199,
+      "learning_rate": 5.703960396039604e-06,
+      "loss": 0.0007,
+      "step": 192
+    },
+    {
+      "epoch": 0.947239263803681,
+      "grad_norm": 0.020626569166779518,
+      "learning_rate": 5.697029702970297e-06,
+      "loss": 0.002,
+      "step": 193
+    },
+    {
+      "epoch": 0.9521472392638037,
+      "grad_norm": 0.009778481908142567,
+      "learning_rate": 5.69009900990099e-06,
+      "loss": 0.0011,
+      "step": 194
+    },
+    {
+      "epoch": 0.9570552147239264,
+      "grad_norm": 0.020891468971967697,
+      "learning_rate": 5.683168316831684e-06,
+      "loss": 0.0017,
+      "step": 195
+    },
+    {
+      "epoch": 0.9619631901840491,
+      "grad_norm": 0.01422948855906725,
+      "learning_rate": 5.676237623762377e-06,
+      "loss": 0.0017,
+      "step": 196
+    },
+    {
+      "epoch": 0.9668711656441717,
+      "grad_norm": 0.3881951868534088,
+      "learning_rate": 5.66930693069307e-06,
+      "loss": 0.0239,
+      "step": 197
+    },
+    {
+      "epoch": 0.9717791411042945,
+      "grad_norm": 0.19655589759349823,
+      "learning_rate": 5.662376237623763e-06,
+      "loss": 0.0166,
+      "step": 198
+    },
+    {
+      "epoch": 0.9766871165644172,
+      "grad_norm": 0.22294123470783234,
+      "learning_rate": 5.655445544554456e-06,
+      "loss": 0.0106,
+      "step": 199
+    },
+    {
+      "epoch": 0.9815950920245399,
+      "grad_norm": 0.13569732010364532,
+      "learning_rate": 5.648514851485149e-06,
+      "loss": 0.0054,
+      "step": 200
+    },
+    {
+      "epoch": 0.9865030674846625,
+      "grad_norm": 0.1216830313205719,
+      "learning_rate": 5.641584158415842e-06,
+      "loss": 0.0075,
+      "step": 201
+    },
+    {
+      "epoch": 0.9914110429447853,
+      "grad_norm": 0.14887885749340057,
+      "learning_rate": 5.634653465346535e-06,
+      "loss": 0.0059,
+      "step": 202
+    },
+    {
+      "epoch": 0.996319018404908,
+      "grad_norm": 0.34408119320869446,
+      "learning_rate": 5.627722772277228e-06,
+      "loss": 0.0052,
+      "step": 203
+    },
+    {
+      "epoch": 1.0012269938650307,
+      "grad_norm": 0.022362129762768745,
+      "learning_rate": 5.620792079207921e-06,
+      "loss": 0.0016,
+      "step": 204
+    },
+    {
+      "epoch": 1.0061349693251533,
+      "grad_norm": 0.2365986853837967,
+      "learning_rate": 5.6138613861386145e-06,
+      "loss": 0.0205,
+      "step": 205
+    },
+    {
+      "epoch": 1.011042944785276,
+      "grad_norm": 0.007400259375572205,
+      "learning_rate": 5.6069306930693075e-06,
+      "loss": 0.0007,
+      "step": 206
+    },
+    {
+      "epoch": 1.0159509202453987,
+      "grad_norm": 0.016991781070828438,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.0019,
+      "step": 207
+    },
+    {
+      "epoch": 1.0208588957055216,
+      "grad_norm": 0.035710081458091736,
+      "learning_rate": 5.593069306930694e-06,
+      "loss": 0.0016,
+      "step": 208
+    },
+    {
+      "epoch": 1.0257668711656442,
+      "grad_norm": 0.16605080664157867,
+      "learning_rate": 5.586138613861387e-06,
+      "loss": 0.0127,
+      "step": 209
+    },
+    {
+      "epoch": 1.030674846625767,
+      "grad_norm": 0.029172230511903763,
+      "learning_rate": 5.57920792079208e-06,
+      "loss": 0.0026,
+      "step": 210
+    },
+    {
+      "epoch": 1.0355828220858896,
+      "grad_norm": 0.012522108852863312,
+      "learning_rate": 5.572277227722773e-06,
+      "loss": 0.0013,
+      "step": 211
+    },
+    {
+      "epoch": 1.0404907975460123,
+      "grad_norm": 0.2723598778247833,
+      "learning_rate": 5.565346534653466e-06,
+      "loss": 0.0244,
+      "step": 212
+    },
+    {
+      "epoch": 1.045398773006135,
+      "grad_norm": 0.10059408843517303,
+      "learning_rate": 5.558415841584159e-06,
+      "loss": 0.0052,
+      "step": 213
+    },
+    {
+      "epoch": 1.0503067484662576,
+      "grad_norm": 0.040718916803598404,
+      "learning_rate": 5.551485148514852e-06,
+      "loss": 0.0027,
+      "step": 214
+    },
+    {
+      "epoch": 1.0552147239263803,
+      "grad_norm": 0.044893424957990646,
+      "learning_rate": 5.544554455445545e-06,
+      "loss": 0.0025,
+      "step": 215
+    },
+    {
+      "epoch": 1.0601226993865032,
+      "grad_norm": 0.0643078088760376,
+      "learning_rate": 5.537623762376238e-06,
+      "loss": 0.0026,
+      "step": 216
+    },
+    {
+      "epoch": 1.0650306748466258,
+      "grad_norm": 0.23729190230369568,
+      "learning_rate": 5.530693069306931e-06,
+      "loss": 0.0073,
+      "step": 217
+    },
+    {
+      "epoch": 1.0699386503067485,
+      "grad_norm": 0.03545796498656273,
+      "learning_rate": 5.5237623762376235e-06,
+      "loss": 0.0043,
+      "step": 218
+    },
+    {
+      "epoch": 1.0748466257668712,
+      "grad_norm": 0.12645600736141205,
+      "learning_rate": 5.5168316831683166e-06,
+      "loss": 0.0168,
+      "step": 219
+    },
+    {
+      "epoch": 1.0797546012269938,
+      "grad_norm": 0.22768345475196838,
+      "learning_rate": 5.50990099009901e-06,
+      "loss": 0.0066,
+      "step": 220
+    },
+    {
+      "epoch": 1.0846625766871165,
+      "grad_norm": 0.008956272155046463,
+      "learning_rate": 5.502970297029703e-06,
+      "loss": 0.002,
+      "step": 221
+    },
+    {
+      "epoch": 1.0895705521472392,
+      "grad_norm": 0.14142072200775146,
+      "learning_rate": 5.496039603960396e-06,
+      "loss": 0.005,
+      "step": 222
+    },
+    {
+      "epoch": 1.0944785276073619,
+      "grad_norm": 0.3319880962371826,
+      "learning_rate": 5.489108910891089e-06,
+      "loss": 0.0116,
+      "step": 223
+    },
+    {
+      "epoch": 1.0993865030674848,
+      "grad_norm": 0.03383160009980202,
+      "learning_rate": 5.482178217821782e-06,
+      "loss": 0.0021,
+      "step": 224
+    },
+    {
+      "epoch": 1.1042944785276074,
+      "grad_norm": 0.03335406631231308,
+      "learning_rate": 5.475247524752475e-06,
+      "loss": 0.0022,
+      "step": 225
+    },
+    {
+      "epoch": 1.10920245398773,
+      "grad_norm": 0.06353459507226944,
+      "learning_rate": 5.468316831683168e-06,
+      "loss": 0.0062,
+      "step": 226
+    },
+    {
+      "epoch": 1.1141104294478528,
+      "grad_norm": 0.2315845489501953,
+      "learning_rate": 5.461386138613861e-06,
+      "loss": 0.0221,
+      "step": 227
+    },
+    {
+      "epoch": 1.1190184049079754,
+      "grad_norm": 0.04464046284556389,
+      "learning_rate": 5.454455445544554e-06,
+      "loss": 0.0023,
+      "step": 228
+    },
+    {
+      "epoch": 1.123926380368098,
+      "grad_norm": 0.49593213200569153,
+      "learning_rate": 5.447524752475247e-06,
+      "loss": 0.0484,
+      "step": 229
+    },
+    {
+      "epoch": 1.1288343558282208,
+      "grad_norm": 0.050679679960012436,
+      "learning_rate": 5.44059405940594e-06,
+      "loss": 0.0021,
+      "step": 230
+    },
+    {
+      "epoch": 1.1337423312883437,
+      "grad_norm": 0.08641895651817322,
+      "learning_rate": 5.433663366336633e-06,
+      "loss": 0.0022,
+      "step": 231
+    },
+    {
+      "epoch": 1.1386503067484663,
+      "grad_norm": 0.046292319893836975,
+      "learning_rate": 5.4267326732673265e-06,
+      "loss": 0.0034,
+      "step": 232
+    },
+    {
+      "epoch": 1.143558282208589,
+      "grad_norm": 0.11266610026359558,
+      "learning_rate": 5.4198019801980195e-06,
+      "loss": 0.0048,
+      "step": 233
+    },
+    {
+      "epoch": 1.1484662576687117,
+      "grad_norm": 0.4683886170387268,
+      "learning_rate": 5.412871287128713e-06,
+      "loss": 0.0095,
+      "step": 234
+    },
+    {
+      "epoch": 1.1533742331288344,
+      "grad_norm": 0.17976729571819305,
+      "learning_rate": 5.405940594059406e-06,
+      "loss": 0.0221,
+      "step": 235
+    },
+    {
+      "epoch": 1.158282208588957,
+      "grad_norm": 0.08795922249555588,
+      "learning_rate": 5.399009900990099e-06,
+      "loss": 0.003,
+      "step": 236
+    },
+    {
+      "epoch": 1.1631901840490797,
+      "grad_norm": 0.012228120118379593,
+      "learning_rate": 5.392079207920792e-06,
+      "loss": 0.0007,
+      "step": 237
+    },
+    {
+      "epoch": 1.1680981595092024,
+      "grad_norm": 0.04390159249305725,
+      "learning_rate": 5.385148514851485e-06,
+      "loss": 0.0019,
+      "step": 238
+    },
+    {
+      "epoch": 1.173006134969325,
+      "grad_norm": 0.03781568259000778,
+      "learning_rate": 5.378217821782178e-06,
+      "loss": 0.002,
+      "step": 239
+    },
+    {
+      "epoch": 1.177914110429448,
+      "grad_norm": 0.18868562579154968,
+      "learning_rate": 5.371287128712871e-06,
+      "loss": 0.0078,
+      "step": 240
+    },
+    {
+      "epoch": 1.1828220858895706,
+      "grad_norm": 0.12037086486816406,
+      "learning_rate": 5.364356435643564e-06,
+      "loss": 0.0047,
+      "step": 241
+    },
+    {
+      "epoch": 1.1877300613496933,
+      "grad_norm": 0.21462522447109222,
+      "learning_rate": 5.357425742574257e-06,
+      "loss": 0.0075,
+      "step": 242
+    },
+    {
+      "epoch": 1.192638036809816,
+      "grad_norm": 0.14189103245735168,
+      "learning_rate": 5.35049504950495e-06,
+      "loss": 0.0109,
+      "step": 243
+    },
+    {
+      "epoch": 1.1975460122699386,
+      "grad_norm": 0.30977487564086914,
+      "learning_rate": 5.343564356435643e-06,
+      "loss": 0.0142,
+      "step": 244
+    },
+    {
+      "epoch": 1.2024539877300613,
+      "grad_norm": 0.03001783974468708,
+      "learning_rate": 5.336633663366336e-06,
+      "loss": 0.0011,
+      "step": 245
+    },
+    {
+      "epoch": 1.207361963190184,
+      "grad_norm": 0.35600486397743225,
+      "learning_rate": 5.329702970297029e-06,
+      "loss": 0.0343,
+      "step": 246
+    },
+    {
+      "epoch": 1.2122699386503069,
+      "grad_norm": 0.023076960816979408,
+      "learning_rate": 5.3227722772277225e-06,
+      "loss": 0.0021,
+      "step": 247
+    },
+    {
+      "epoch": 1.2171779141104295,
+      "grad_norm": 0.23361201584339142,
+      "learning_rate": 5.3158415841584156e-06,
+      "loss": 0.0068,
+      "step": 248
+    },
+    {
+      "epoch": 1.2220858895705522,
+      "grad_norm": 0.21508128941059113,
+      "learning_rate": 5.308910891089109e-06,
+      "loss": 0.0247,
+      "step": 249
+    },
+    {
+      "epoch": 1.2269938650306749,
+      "grad_norm": 0.1582375317811966,
+      "learning_rate": 5.301980198019802e-06,
+      "loss": 0.0037,
+      "step": 250
+    },
+    {
+      "epoch": 1.2319018404907975,
+      "grad_norm": 0.037315189838409424,
+      "learning_rate": 5.295049504950495e-06,
+      "loss": 0.0014,
+      "step": 251
+    },
+    {
+      "epoch": 1.2368098159509202,
+      "grad_norm": 0.07501725107431412,
+      "learning_rate": 5.288118811881188e-06,
+      "loss": 0.0037,
+      "step": 252
+    },
+    {
+      "epoch": 1.2417177914110429,
+      "grad_norm": 0.0727037638425827,
+      "learning_rate": 5.281188118811881e-06,
+      "loss": 0.0028,
+      "step": 253
+    },
+    {
+      "epoch": 1.2466257668711656,
+      "grad_norm": 0.04310867562890053,
+      "learning_rate": 5.274257425742574e-06,
+      "loss": 0.0039,
+      "step": 254
+    },
+    {
+      "epoch": 1.2515337423312882,
+      "grad_norm": 0.042814526706933975,
+      "learning_rate": 5.267326732673267e-06,
+      "loss": 0.0025,
+      "step": 255
+    },
+    {
+      "epoch": 1.2564417177914111,
+      "grad_norm": 0.11775332689285278,
+      "learning_rate": 5.26039603960396e-06,
+      "loss": 0.0127,
+      "step": 256
+    },
+    {
+      "epoch": 1.2613496932515338,
+      "grad_norm": 0.040224045515060425,
+      "learning_rate": 5.253465346534653e-06,
+      "loss": 0.0017,
+      "step": 257
+    },
+    {
+      "epoch": 1.2662576687116565,
+      "grad_norm": 0.011041068471968174,
+      "learning_rate": 5.246534653465346e-06,
+      "loss": 0.0011,
+      "step": 258
+    },
+    {
+      "epoch": 1.2711656441717791,
+      "grad_norm": 0.019592339172959328,
+      "learning_rate": 5.239603960396039e-06,
+      "loss": 0.0013,
+      "step": 259
+    },
+    {
+      "epoch": 1.2760736196319018,
+      "grad_norm": 0.0652371346950531,
+      "learning_rate": 5.232673267326732e-06,
+      "loss": 0.0026,
+      "step": 260
+    },
+    {
+      "epoch": 1.2809815950920245,
+      "grad_norm": 0.5446717143058777,
+      "learning_rate": 5.2257425742574254e-06,
+      "loss": 0.0305,
+      "step": 261
+    },
+    {
+      "epoch": 1.2858895705521474,
+      "grad_norm": 0.02951255440711975,
+      "learning_rate": 5.2188118811881185e-06,
+      "loss": 0.001,
+      "step": 262
+    },
+    {
+      "epoch": 1.29079754601227,
+      "grad_norm": 0.1153414323925972,
+      "learning_rate": 5.211881188118812e-06,
+      "loss": 0.0088,
+      "step": 263
+    },
+    {
+      "epoch": 1.2957055214723927,
+      "grad_norm": 0.17192097008228302,
+      "learning_rate": 5.204950495049505e-06,
+      "loss": 0.0043,
+      "step": 264
+    },
+    {
+      "epoch": 1.3006134969325154,
+      "grad_norm": 0.04299500212073326,
+      "learning_rate": 5.198019801980198e-06,
+      "loss": 0.0042,
+      "step": 265
+    },
+    {
+      "epoch": 1.305521472392638,
+      "grad_norm": 0.02218775264918804,
+      "learning_rate": 5.191089108910891e-06,
+      "loss": 0.001,
+      "step": 266
+    },
+    {
+      "epoch": 1.3104294478527607,
+      "grad_norm": 0.7823259830474854,
+      "learning_rate": 5.184158415841584e-06,
+      "loss": 0.0556,
+      "step": 267
+    },
+    {
+      "epoch": 1.3153374233128834,
+      "grad_norm": 0.09460531175136566,
+      "learning_rate": 5.177227722772277e-06,
+      "loss": 0.0091,
+      "step": 268
+    },
+    {
+      "epoch": 1.320245398773006,
+      "grad_norm": 0.04762015491724014,
+      "learning_rate": 5.17029702970297e-06,
+      "loss": 0.0039,
+      "step": 269
+    },
+    {
+      "epoch": 1.3251533742331287,
+      "grad_norm": 0.033739686012268066,
+      "learning_rate": 5.163366336633663e-06,
+      "loss": 0.0017,
+      "step": 270
+    },
+    {
+      "epoch": 1.3300613496932514,
+      "grad_norm": 0.06530934572219849,
+      "learning_rate": 5.156435643564356e-06,
+      "loss": 0.0047,
+      "step": 271
+    },
+    {
+      "epoch": 1.3349693251533743,
+      "grad_norm": 0.014586723409593105,
+      "learning_rate": 5.149504950495049e-06,
+      "loss": 0.001,
+      "step": 272
+    },
+    {
+      "epoch": 1.339877300613497,
+      "grad_norm": 0.05435250699520111,
+      "learning_rate": 5.142574257425742e-06,
+      "loss": 0.0028,
+      "step": 273
+    },
+    {
+      "epoch": 1.3447852760736196,
+      "grad_norm": 0.1281612366437912,
+      "learning_rate": 5.135643564356435e-06,
+      "loss": 0.0026,
+      "step": 274
+    },
+    {
+      "epoch": 1.3496932515337423,
+      "grad_norm": 0.18499918282032013,
+      "learning_rate": 5.128712871287128e-06,
+      "loss": 0.0057,
+      "step": 275
+    },
+    {
+      "epoch": 1.354601226993865,
+      "grad_norm": 0.07541365176439285,
+      "learning_rate": 5.1217821782178215e-06,
+      "loss": 0.003,
+      "step": 276
+    },
+    {
+      "epoch": 1.3595092024539877,
+      "grad_norm": 0.061677657067775726,
+      "learning_rate": 5.1148514851485145e-06,
+      "loss": 0.002,
+      "step": 277
+    },
+    {
+      "epoch": 1.3644171779141105,
+      "grad_norm": 0.16630059480667114,
+      "learning_rate": 5.107920792079208e-06,
+      "loss": 0.0138,
+      "step": 278
+    },
+    {
+      "epoch": 1.3693251533742332,
+      "grad_norm": 0.13260015845298767,
+      "learning_rate": 5.100990099009901e-06,
+      "loss": 0.0015,
+      "step": 279
+    },
+    {
+      "epoch": 1.3742331288343559,
+      "grad_norm": 0.04029810056090355,
+      "learning_rate": 5.094059405940594e-06,
+      "loss": 0.0022,
+      "step": 280
+    },
+    {
+      "epoch": 1.3791411042944786,
+      "grad_norm": 0.07820600271224976,
+      "learning_rate": 5.087128712871287e-06,
+      "loss": 0.0012,
+      "step": 281
+    },
+    {
+      "epoch": 1.3840490797546012,
+      "grad_norm": 0.11421211808919907,
+      "learning_rate": 5.08019801980198e-06,
+      "loss": 0.0036,
+      "step": 282
+    },
+    {
+      "epoch": 1.388957055214724,
+      "grad_norm": 0.02426535077393055,
+      "learning_rate": 5.073267326732673e-06,
+      "loss": 0.0015,
+      "step": 283
+    },
+    {
+      "epoch": 1.3938650306748466,
+      "grad_norm": 0.00859303679317236,
+      "learning_rate": 5.066336633663366e-06,
+      "loss": 0.001,
+      "step": 284
+    },
+    {
+      "epoch": 1.3987730061349692,
+      "grad_norm": 0.03431880474090576,
+      "learning_rate": 5.059405940594059e-06,
+      "loss": 0.0021,
+      "step": 285
+    },
+    {
+      "epoch": 1.403680981595092,
+      "grad_norm": 0.013164886273443699,
+      "learning_rate": 5.052475247524752e-06,
+      "loss": 0.0009,
+      "step": 286
+    },
+    {
+      "epoch": 1.4085889570552146,
+      "grad_norm": 0.1018059104681015,
+      "learning_rate": 5.045544554455445e-06,
+      "loss": 0.003,
+      "step": 287
+    },
+    {
+      "epoch": 1.4134969325153375,
+      "grad_norm": 0.3569484055042267,
+      "learning_rate": 5.038613861386138e-06,
+      "loss": 0.0193,
+      "step": 288
+    },
+    {
+      "epoch": 1.4184049079754601,
+      "grad_norm": 0.03560711443424225,
+      "learning_rate": 5.031683168316831e-06,
+      "loss": 0.0041,
+      "step": 289
+    },
+    {
+      "epoch": 1.4233128834355828,
+      "grad_norm": 0.049912016838788986,
+      "learning_rate": 5.0247524752475244e-06,
+      "loss": 0.0019,
+      "step": 290
+    },
+    {
+      "epoch": 1.4282208588957055,
+      "grad_norm": 0.00541495019569993,
+      "learning_rate": 5.0178217821782175e-06,
+      "loss": 0.0008,
+      "step": 291
+    },
+    {
+      "epoch": 1.4331288343558282,
+      "grad_norm": 0.188345804810524,
+      "learning_rate": 5.0108910891089106e-06,
+      "loss": 0.0138,
+      "step": 292
+    },
+    {
+      "epoch": 1.438036809815951,
+      "grad_norm": 1.1841977834701538,
+      "learning_rate": 5.003960396039604e-06,
+      "loss": 0.0285,
+      "step": 293
+    },
+    {
+      "epoch": 1.4429447852760737,
+      "grad_norm": 0.01752946898341179,
+      "learning_rate": 4.997029702970297e-06,
+      "loss": 0.003,
+      "step": 294
+    },
+    {
+      "epoch": 1.4478527607361964,
+      "grad_norm": 0.009354379959404469,
+      "learning_rate": 4.99009900990099e-06,
+      "loss": 0.001,
+      "step": 295
+    },
+    {
+      "epoch": 1.452760736196319,
+      "grad_norm": 0.5457541346549988,
+      "learning_rate": 4.983168316831683e-06,
+      "loss": 0.0028,
+      "step": 296
+    },
+    {
+      "epoch": 1.4576687116564417,
+      "grad_norm": 0.012560038827359676,
+      "learning_rate": 4.976237623762376e-06,
+      "loss": 0.0014,
+      "step": 297
+    },
+    {
+      "epoch": 1.4625766871165644,
+      "grad_norm": 0.08460962772369385,
+      "learning_rate": 4.969306930693069e-06,
+      "loss": 0.0054,
+      "step": 298
+    },
+    {
+      "epoch": 1.467484662576687,
+      "grad_norm": 0.02863544225692749,
+      "learning_rate": 4.962376237623762e-06,
+      "loss": 0.0019,
+      "step": 299
+    },
+    {
+      "epoch": 1.4723926380368098,
+      "grad_norm": 0.12013754993677139,
+      "learning_rate": 4.955445544554455e-06,
+      "loss": 0.0108,
+      "step": 300
+    },
+    {
+      "epoch": 1.4773006134969324,
+      "grad_norm": 0.06849315017461777,
+      "learning_rate": 4.948514851485148e-06,
+      "loss": 0.0023,
+      "step": 301
+    },
+    {
+      "epoch": 1.482208588957055,
+      "grad_norm": 0.07875117659568787,
+      "learning_rate": 4.941584158415841e-06,
+      "loss": 0.003,
+      "step": 302
+    },
+    {
+      "epoch": 1.487116564417178,
+      "grad_norm": 0.03466253727674484,
+      "learning_rate": 4.934653465346534e-06,
+      "loss": 0.0023,
+      "step": 303
+    },
+    {
+      "epoch": 1.4920245398773007,
+      "grad_norm": 0.02622813917696476,
+      "learning_rate": 4.927722772277227e-06,
+      "loss": 0.0019,
+      "step": 304
+    },
+    {
+      "epoch": 1.4969325153374233,
+      "grad_norm": 0.08774946630001068,
+      "learning_rate": 4.9207920792079205e-06,
+      "loss": 0.0062,
+      "step": 305
+    },
+    {
+      "epoch": 1.501840490797546,
+      "grad_norm": 0.038571231067180634,
+      "learning_rate": 4.9138613861386135e-06,
+      "loss": 0.002,
+      "step": 306
+    },
+    {
+      "epoch": 1.5067484662576687,
+      "grad_norm": 0.058768339455127716,
+      "learning_rate": 4.906930693069307e-06,
+      "loss": 0.0024,
+      "step": 307
+    },
+    {
+      "epoch": 1.5116564417177916,
+      "grad_norm": 0.01673268899321556,
+      "learning_rate": 4.9e-06,
+      "loss": 0.0016,
+      "step": 308
+    },
+    {
+      "epoch": 1.5165644171779142,
+      "grad_norm": 0.1858752965927124,
+      "learning_rate": 4.893069306930693e-06,
+      "loss": 0.0042,
+      "step": 309
+    },
+    {
+      "epoch": 1.521472392638037,
+      "grad_norm": 0.017566069960594177,
+      "learning_rate": 4.886138613861386e-06,
+      "loss": 0.0012,
+      "step": 310
+    },
+    {
+      "epoch": 1.5263803680981596,
+      "grad_norm": 0.021138716489076614,
+      "learning_rate": 4.879207920792079e-06,
+      "loss": 0.0018,
+      "step": 311
+    },
+    {
+      "epoch": 1.5312883435582823,
+      "grad_norm": 0.005446423310786486,
+      "learning_rate": 4.872277227722772e-06,
+      "loss": 0.0012,
+      "step": 312
+    },
+    {
+      "epoch": 1.536196319018405,
+      "grad_norm": 0.034122321754693985,
+      "learning_rate": 4.865346534653465e-06,
+      "loss": 0.0029,
+      "step": 313
+    },
+    {
+      "epoch": 1.5411042944785276,
+      "grad_norm": 0.015392904169857502,
+      "learning_rate": 4.858415841584158e-06,
+      "loss": 0.001,
+      "step": 314
+    },
+    {
+      "epoch": 1.5460122699386503,
+      "grad_norm": 0.09795702993869781,
+      "learning_rate": 4.851485148514851e-06,
+      "loss": 0.006,
+      "step": 315
+    },
+    {
+      "epoch": 1.550920245398773,
+      "grad_norm": 0.18449151515960693,
+      "learning_rate": 4.844554455445544e-06,
+      "loss": 0.0543,
+      "step": 316
+    },
+    {
+      "epoch": 1.5558282208588956,
+      "grad_norm": 0.3047311007976532,
+      "learning_rate": 4.837623762376237e-06,
+      "loss": 0.0148,
+      "step": 317
+    },
+    {
+      "epoch": 1.5607361963190183,
+      "grad_norm": 0.3502406179904938,
+      "learning_rate": 4.83069306930693e-06,
+      "loss": 0.0083,
+      "step": 318
+    },
+    {
+      "epoch": 1.565644171779141,
+      "grad_norm": 0.08116799592971802,
+      "learning_rate": 4.8237623762376234e-06,
+      "loss": 0.011,
+      "step": 319
+    },
+    {
+      "epoch": 1.5705521472392638,
+      "grad_norm": 6.128328323364258,
+      "learning_rate": 4.8168316831683165e-06,
+      "loss": 0.0234,
+      "step": 320
+    },
+    {
+      "epoch": 1.5754601226993865,
+      "grad_norm": 0.31991851329803467,
+      "learning_rate": 4.80990099009901e-06,
+      "loss": 0.0161,
+      "step": 321
+    },
+    {
+      "epoch": 1.5803680981595092,
+      "grad_norm": 0.08932381123304367,
+      "learning_rate": 4.8029702970297035e-06,
+      "loss": 0.0024,
+      "step": 322
+    },
+    {
+      "epoch": 1.5852760736196319,
+      "grad_norm": 0.03641815483570099,
+      "learning_rate": 4.7960396039603965e-06,
+      "loss": 0.0013,
+      "step": 323
+    },
+    {
+      "epoch": 1.5901840490797547,
+      "grad_norm": 0.12057910859584808,
+      "learning_rate": 4.78910891089109e-06,
+      "loss": 0.0045,
+      "step": 324
+    },
+    {
+      "epoch": 1.5950920245398774,
+      "grad_norm": 0.022239159792661667,
+      "learning_rate": 4.782178217821783e-06,
+      "loss": 0.002,
+      "step": 325
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.02001642808318138,
+      "learning_rate": 4.775247524752476e-06,
+      "loss": 0.0011,
+      "step": 326
+    },
+    {
+      "epoch": 1.6049079754601228,
+      "grad_norm": 0.0681408941745758,
+      "learning_rate": 4.768316831683169e-06,
+      "loss": 0.0034,
+      "step": 327
+    },
+    {
+      "epoch": 1.6098159509202454,
+      "grad_norm": 0.07200402021408081,
+      "learning_rate": 4.761386138613862e-06,
+      "loss": 0.0023,
+      "step": 328
+    },
+    {
+      "epoch": 1.614723926380368,
+      "grad_norm": 0.11577661335468292,
+      "learning_rate": 4.754455445544555e-06,
+      "loss": 0.0122,
+      "step": 329
+    },
+    {
+      "epoch": 1.6196319018404908,
+      "grad_norm": 0.12311868369579315,
+      "learning_rate": 4.747524752475248e-06,
+      "loss": 0.0148,
+      "step": 330
+    },
+    {
+      "epoch": 1.6245398773006134,
+      "grad_norm": 0.07884922623634338,
+      "learning_rate": 4.740594059405941e-06,
+      "loss": 0.0026,
+      "step": 331
+    },
+    {
+      "epoch": 1.6294478527607361,
+      "grad_norm": 0.013971218839287758,
+      "learning_rate": 4.733663366336634e-06,
+      "loss": 0.0014,
+      "step": 332
+    },
+    {
+      "epoch": 1.6343558282208588,
+      "grad_norm": 0.17173822224140167,
+      "learning_rate": 4.726732673267327e-06,
+      "loss": 0.0041,
+      "step": 333
+    },
+    {
+      "epoch": 1.6392638036809815,
+      "grad_norm": 0.24089990556240082,
+      "learning_rate": 4.71980198019802e-06,
+      "loss": 0.0114,
+      "step": 334
+    },
+    {
+      "epoch": 1.6441717791411041,
+      "grad_norm": 0.026243751868605614,
+      "learning_rate": 4.712871287128713e-06,
+      "loss": 0.0011,
+      "step": 335
+    },
+    {
+      "epoch": 1.649079754601227,
+      "grad_norm": 0.08318141102790833,
+      "learning_rate": 4.7059405940594064e-06,
+      "loss": 0.0034,
+      "step": 336
+    },
+    {
+      "epoch": 1.6539877300613497,
+      "grad_norm": 0.010400490835309029,
+      "learning_rate": 4.6990099009900995e-06,
+      "loss": 0.0022,
+      "step": 337
+    },
+    {
+      "epoch": 1.6588957055214724,
+      "grad_norm": 0.044309504330158234,
+      "learning_rate": 4.692079207920793e-06,
+      "loss": 0.004,
+      "step": 338
+    },
+    {
+      "epoch": 1.6638036809815953,
+      "grad_norm": 0.02144954912364483,
+      "learning_rate": 4.685148514851486e-06,
+      "loss": 0.0018,
+      "step": 339
+    },
+    {
+      "epoch": 1.668711656441718,
+      "grad_norm": 0.0410357341170311,
+      "learning_rate": 4.678217821782179e-06,
+      "loss": 0.002,
+      "step": 340
+    },
+    {
+      "epoch": 1.6736196319018406,
+      "grad_norm": 0.02522197738289833,
+      "learning_rate": 4.671287128712872e-06,
+      "loss": 0.0014,
+      "step": 341
+    },
+    {
+      "epoch": 1.6785276073619633,
+      "grad_norm": 0.008343094028532505,
+      "learning_rate": 4.664356435643565e-06,
+      "loss": 0.0016,
+      "step": 342
+    },
+    {
+      "epoch": 1.683435582822086,
+      "grad_norm": 0.19290384650230408,
+      "learning_rate": 4.657425742574258e-06,
+      "loss": 0.0063,
+      "step": 343
+    },
+    {
+      "epoch": 1.6883435582822086,
+      "grad_norm": 0.038624536246061325,
+      "learning_rate": 4.650495049504951e-06,
+      "loss": 0.0016,
+      "step": 344
+    },
+    {
+      "epoch": 1.6932515337423313,
+      "grad_norm": 0.21920740604400635,
+      "learning_rate": 4.643564356435644e-06,
+      "loss": 0.0105,
+      "step": 345
+    },
+    {
+      "epoch": 1.698159509202454,
+      "grad_norm": 0.057425156235694885,
+      "learning_rate": 4.636633663366337e-06,
+      "loss": 0.0033,
+      "step": 346
+    },
+    {
+      "epoch": 1.7030674846625766,
+      "grad_norm": 0.024981388822197914,
+      "learning_rate": 4.62970297029703e-06,
+      "loss": 0.0025,
+      "step": 347
+    },
+    {
+      "epoch": 1.7079754601226993,
+      "grad_norm": 0.020687798038125038,
+      "learning_rate": 4.622772277227723e-06,
+      "loss": 0.0022,
+      "step": 348
+    },
+    {
+      "epoch": 1.712883435582822,
+      "grad_norm": 0.028213948011398315,
+      "learning_rate": 4.615841584158416e-06,
+      "loss": 0.0014,
+      "step": 349
+    },
+    {
+      "epoch": 1.7177914110429446,
+      "grad_norm": 0.05201176926493645,
+      "learning_rate": 4.608910891089109e-06,
+      "loss": 0.0068,
+      "step": 350
+    },
+    {
+      "epoch": 1.7226993865030675,
+      "grad_norm": 0.057419948279857635,
+      "learning_rate": 4.6019801980198025e-06,
+      "loss": 0.0027,
+      "step": 351
+    },
+    {
+      "epoch": 1.7276073619631902,
+      "grad_norm": 0.026078782975673676,
+      "learning_rate": 4.5950495049504955e-06,
+      "loss": 0.0017,
+      "step": 352
+    },
+    {
+      "epoch": 1.7325153374233129,
+      "grad_norm": 0.09312699735164642,
+      "learning_rate": 4.588118811881189e-06,
+      "loss": 0.0048,
+      "step": 353
+    },
+    {
+      "epoch": 1.7374233128834355,
+      "grad_norm": 0.027963656932115555,
+      "learning_rate": 4.581188118811882e-06,
+      "loss": 0.0018,
+      "step": 354
+    },
+    {
+      "epoch": 1.7423312883435584,
+      "grad_norm": 1.0171724557876587,
+      "learning_rate": 4.574257425742575e-06,
+      "loss": 0.044,
+      "step": 355
+    },
+    {
+      "epoch": 1.747239263803681,
+      "grad_norm": 0.0031565092504024506,
+      "learning_rate": 4.567326732673268e-06,
+      "loss": 0.0003,
+      "step": 356
+    },
+    {
+      "epoch": 1.7521472392638038,
+      "grad_norm": 0.023803675547242165,
+      "learning_rate": 4.560396039603961e-06,
+      "loss": 0.003,
+      "step": 357
+    },
+    {
+      "epoch": 1.7570552147239265,
+      "grad_norm": 0.11927448213100433,
+      "learning_rate": 4.553465346534654e-06,
+      "loss": 0.0046,
+      "step": 358
+    },
+    {
+      "epoch": 1.7619631901840491,
+      "grad_norm": 0.20977821946144104,
+      "learning_rate": 4.546534653465347e-06,
+      "loss": 0.0287,
+      "step": 359
+    },
+    {
+      "epoch": 1.7668711656441718,
+      "grad_norm": 0.024297788739204407,
+      "learning_rate": 4.53960396039604e-06,
+      "loss": 0.0016,
+      "step": 360
+    },
+    {
+      "epoch": 1.7717791411042945,
+      "grad_norm": 0.0834914818406105,
+      "learning_rate": 4.532673267326733e-06,
+      "loss": 0.0032,
+      "step": 361
+    },
+    {
+      "epoch": 1.7766871165644171,
+      "grad_norm": 0.07108546048402786,
+      "learning_rate": 4.525742574257426e-06,
+      "loss": 0.0021,
+      "step": 362
+    },
+    {
+      "epoch": 1.7815950920245398,
+      "grad_norm": 0.037000950425863266,
+      "learning_rate": 4.518811881188119e-06,
+      "loss": 0.0019,
+      "step": 363
+    },
+    {
+      "epoch": 1.7865030674846625,
+      "grad_norm": 0.050357453525066376,
+      "learning_rate": 4.5118811881188115e-06,
+      "loss": 0.0017,
+      "step": 364
+    },
+    {
+      "epoch": 1.7914110429447851,
+      "grad_norm": 0.15029194951057434,
+      "learning_rate": 4.504950495049505e-06,
+      "loss": 0.0054,
+      "step": 365
+    },
+    {
+      "epoch": 1.7963190184049078,
+      "grad_norm": 0.10820876061916351,
+      "learning_rate": 4.498019801980198e-06,
+      "loss": 0.0029,
+      "step": 366
+    },
+    {
+      "epoch": 1.8012269938650307,
+      "grad_norm": 0.014579812064766884,
+      "learning_rate": 4.491089108910891e-06,
+      "loss": 0.0014,
+      "step": 367
+    },
+    {
+      "epoch": 1.8061349693251534,
+      "grad_norm": 0.022982638329267502,
+      "learning_rate": 4.484158415841584e-06,
+      "loss": 0.0012,
+      "step": 368
+    },
+    {
+      "epoch": 1.811042944785276,
+      "grad_norm": 0.12965136766433716,
+      "learning_rate": 4.477227722772277e-06,
+      "loss": 0.0059,
+      "step": 369
+    },
+    {
+      "epoch": 1.8159509202453987,
+      "grad_norm": 0.2906913161277771,
+      "learning_rate": 4.47029702970297e-06,
+      "loss": 0.0061,
+      "step": 370
+    },
+    {
+      "epoch": 1.8208588957055216,
+      "grad_norm": 0.04156769439578056,
+      "learning_rate": 4.463366336633663e-06,
+      "loss": 0.0031,
+      "step": 371
+    },
+    {
+      "epoch": 1.8257668711656443,
+      "grad_norm": 0.07649008929729462,
+      "learning_rate": 4.456435643564356e-06,
+      "loss": 0.0027,
+      "step": 372
+    },
+    {
+      "epoch": 1.830674846625767,
+      "grad_norm": 0.11019697785377502,
+      "learning_rate": 4.449504950495049e-06,
+      "loss": 0.0023,
+      "step": 373
+    },
+    {
+      "epoch": 1.8355828220858896,
+      "grad_norm": 0.03163640573620796,
+      "learning_rate": 4.442574257425742e-06,
+      "loss": 0.0024,
+      "step": 374
+    },
+    {
+      "epoch": 1.8404907975460123,
+      "grad_norm": 0.2059604823589325,
+      "learning_rate": 4.435643564356435e-06,
+      "loss": 0.0154,
+      "step": 375
+    },
+    {
+      "epoch": 1.845398773006135,
+      "grad_norm": 0.00596796628087759,
+      "learning_rate": 4.428712871287128e-06,
+      "loss": 0.0003,
+      "step": 376
+    },
+    {
+      "epoch": 1.8503067484662576,
+      "grad_norm": 0.1259843111038208,
+      "learning_rate": 4.421782178217821e-06,
+      "loss": 0.002,
+      "step": 377
+    },
+    {
+      "epoch": 1.8552147239263803,
+      "grad_norm": 0.051318589597940445,
+      "learning_rate": 4.4148514851485145e-06,
+      "loss": 0.0054,
+      "step": 378
+    },
+    {
+      "epoch": 1.860122699386503,
+      "grad_norm": 0.022389305755496025,
+      "learning_rate": 4.4079207920792075e-06,
+      "loss": 0.0011,
+      "step": 379
+    },
+    {
+      "epoch": 1.8650306748466257,
+      "grad_norm": 0.03204929828643799,
+      "learning_rate": 4.400990099009901e-06,
+      "loss": 0.002,
+      "step": 380
+    },
+    {
+      "epoch": 1.8699386503067483,
+      "grad_norm": 0.14024299383163452,
+      "learning_rate": 4.394059405940594e-06,
+      "loss": 0.0044,
+      "step": 381
+    },
+    {
+      "epoch": 1.874846625766871,
+      "grad_norm": 0.1390925794839859,
+      "learning_rate": 4.387128712871287e-06,
+      "loss": 0.0125,
+      "step": 382
+    },
+    {
+      "epoch": 1.879754601226994,
+      "grad_norm": 0.19391588866710663,
+      "learning_rate": 4.38019801980198e-06,
+      "loss": 0.0053,
+      "step": 383
+    },
+    {
+      "epoch": 1.8846625766871166,
+      "grad_norm": 0.059033554047346115,
+      "learning_rate": 4.373267326732673e-06,
+      "loss": 0.0036,
+      "step": 384
+    },
+    {
+      "epoch": 1.8895705521472392,
+      "grad_norm": 0.03923160582780838,
+      "learning_rate": 4.366336633663366e-06,
+      "loss": 0.0021,
+      "step": 385
+    },
+    {
+      "epoch": 1.8944785276073621,
+      "grad_norm": 0.2464819848537445,
+      "learning_rate": 4.359405940594059e-06,
+      "loss": 0.0157,
+      "step": 386
+    },
+    {
+      "epoch": 1.8993865030674848,
+      "grad_norm": 0.02465728111565113,
+      "learning_rate": 4.352475247524752e-06,
+      "loss": 0.002,
+      "step": 387
+    },
+    {
+      "epoch": 1.9042944785276075,
+      "grad_norm": 0.017552992329001427,
+      "learning_rate": 4.345544554455445e-06,
+      "loss": 0.0013,
+      "step": 388
+    },
+    {
+      "epoch": 1.9092024539877301,
+      "grad_norm": 0.07727736979722977,
+      "learning_rate": 4.338613861386138e-06,
+      "loss": 0.004,
+      "step": 389
+    },
+    {
+      "epoch": 1.9141104294478528,
+      "grad_norm": 0.0594131238758564,
+      "learning_rate": 4.331683168316831e-06,
+      "loss": 0.0021,
+      "step": 390
+    },
+    {
+      "epoch": 1.9190184049079755,
+      "grad_norm": 0.03132615610957146,
+      "learning_rate": 4.324752475247524e-06,
+      "loss": 0.0057,
+      "step": 391
+    },
+    {
+      "epoch": 1.9239263803680982,
+      "grad_norm": 0.02295020781457424,
+      "learning_rate": 4.3178217821782174e-06,
+      "loss": 0.0014,
+      "step": 392
+    },
+    {
+      "epoch": 1.9288343558282208,
+      "grad_norm": 0.04230022802948952,
+      "learning_rate": 4.3108910891089105e-06,
+      "loss": 0.0035,
+      "step": 393
+    },
+    {
+      "epoch": 1.9337423312883435,
+      "grad_norm": 0.012272198684513569,
+      "learning_rate": 4.3039603960396036e-06,
+      "loss": 0.0027,
+      "step": 394
+    },
+    {
+      "epoch": 1.9386503067484662,
+      "grad_norm": 0.017140038311481476,
+      "learning_rate": 4.297029702970297e-06,
+      "loss": 0.0015,
+      "step": 395
+    },
+    {
+      "epoch": 1.9435582822085888,
+      "grad_norm": 0.011837287805974483,
+      "learning_rate": 4.29009900990099e-06,
+      "loss": 0.0017,
+      "step": 396
+    },
+    {
+      "epoch": 1.9484662576687115,
+      "grad_norm": 0.03815172612667084,
+      "learning_rate": 4.283168316831683e-06,
+      "loss": 0.0036,
+      "step": 397
+    },
+    {
+      "epoch": 1.9533742331288344,
+      "grad_norm": 0.029975654557347298,
+      "learning_rate": 4.276237623762376e-06,
+      "loss": 0.0017,
+      "step": 398
+    },
+    {
+      "epoch": 1.958282208588957,
+      "grad_norm": 0.09029775857925415,
+      "learning_rate": 4.269306930693069e-06,
+      "loss": 0.0034,
+      "step": 399
+    },
+    {
+      "epoch": 1.9631901840490797,
+      "grad_norm": 0.013478988781571388,
+      "learning_rate": 4.262376237623762e-06,
+      "loss": 0.0012,
+      "step": 400
+    },
+    {
+      "epoch": 1.9680981595092024,
+      "grad_norm": 0.14020872116088867,
+      "learning_rate": 4.255445544554455e-06,
+      "loss": 0.0036,
+      "step": 401
+    },
+    {
+      "epoch": 1.9730061349693253,
+      "grad_norm": 0.08427014946937561,
+      "learning_rate": 4.248514851485148e-06,
+      "loss": 0.0032,
+      "step": 402
+    },
+    {
+      "epoch": 1.977914110429448,
+      "grad_norm": 0.02368674799799919,
+      "learning_rate": 4.241584158415841e-06,
+      "loss": 0.0018,
+      "step": 403
+    },
+    {
+      "epoch": 1.9828220858895707,
+      "grad_norm": 0.17322583496570587,
+      "learning_rate": 4.234653465346534e-06,
+      "loss": 0.0075,
+      "step": 404
+    },
+    {
+      "epoch": 1.9877300613496933,
+      "grad_norm": 0.2020941525697708,
+      "learning_rate": 4.227722772277227e-06,
+      "loss": 0.0056,
+      "step": 405
+    },
+    {
+      "epoch": 1.992638036809816,
+      "grad_norm": 0.01503934245556593,
+      "learning_rate": 4.22079207920792e-06,
+      "loss": 0.0014,
+      "step": 406
+    },
+    {
+      "epoch": 1.9975460122699387,
+      "grad_norm": 0.08337133377790451,
+      "learning_rate": 4.2138613861386135e-06,
+      "loss": 0.0022,
+      "step": 407
+    },
+    {
+      "epoch": 2.0024539877300613,
+      "grad_norm": 0.03418401628732681,
+      "learning_rate": 4.2069306930693065e-06,
+      "loss": 0.0014,
+      "step": 408
+    },
+    {
+      "epoch": 2.007361963190184,
+      "grad_norm": 0.15800650417804718,
+      "learning_rate": 4.2e-06,
+      "loss": 0.0152,
+      "step": 409
+    },
+    {
+      "epoch": 2.0122699386503067,
+      "grad_norm": 0.07395298033952713,
+      "learning_rate": 4.193069306930693e-06,
+      "loss": 0.0027,
+      "step": 410
+    },
+    {
+      "epoch": 2.0171779141104293,
+      "grad_norm": 0.013158326968550682,
+      "learning_rate": 4.186138613861386e-06,
+      "loss": 0.0015,
+      "step": 411
+    },
+    {
+      "epoch": 2.022085889570552,
+      "grad_norm": 0.0122208371758461,
+      "learning_rate": 4.179207920792079e-06,
+      "loss": 0.0015,
+      "step": 412
+    },
+    {
+      "epoch": 2.0269938650306747,
+      "grad_norm": 0.04391651973128319,
+      "learning_rate": 4.172277227722772e-06,
+      "loss": 0.0028,
+      "step": 413
+    },
+    {
+      "epoch": 2.0319018404907974,
+      "grad_norm": 0.007677063811570406,
+      "learning_rate": 4.165346534653465e-06,
+      "loss": 0.0004,
+      "step": 414
+    },
+    {
+      "epoch": 2.03680981595092,
+      "grad_norm": 0.01934950426220894,
+      "learning_rate": 4.158415841584158e-06,
+      "loss": 0.0013,
+      "step": 415
+    },
+    {
+      "epoch": 2.041717791411043,
+      "grad_norm": 0.004791139159351587,
+      "learning_rate": 4.151485148514851e-06,
+      "loss": 0.0005,
+      "step": 416
+    },
+    {
+      "epoch": 2.046625766871166,
+      "grad_norm": 0.03984725847840309,
+      "learning_rate": 4.144554455445544e-06,
+      "loss": 0.0015,
+      "step": 417
+    },
+    {
+      "epoch": 2.0515337423312885,
+      "grad_norm": 0.6277703642845154,
+      "learning_rate": 4.137623762376237e-06,
+      "loss": 0.0269,
+      "step": 418
+    },
+    {
+      "epoch": 2.056441717791411,
+      "grad_norm": 0.1700657159090042,
+      "learning_rate": 4.13069306930693e-06,
+      "loss": 0.0196,
+      "step": 419
+    },
+    {
+      "epoch": 2.061349693251534,
+      "grad_norm": 0.045236505568027496,
+      "learning_rate": 4.123762376237623e-06,
+      "loss": 0.003,
+      "step": 420
+    },
+    {
+      "epoch": 2.0662576687116565,
+      "grad_norm": 0.034617915749549866,
+      "learning_rate": 4.116831683168316e-06,
+      "loss": 0.0045,
+      "step": 421
+    },
+    {
+      "epoch": 2.071165644171779,
+      "grad_norm": 0.022638553753495216,
+      "learning_rate": 4.1099009900990095e-06,
+      "loss": 0.0014,
+      "step": 422
+    },
+    {
+      "epoch": 2.076073619631902,
+      "grad_norm": 0.011074609123170376,
+      "learning_rate": 4.1029702970297026e-06,
+      "loss": 0.0006,
+      "step": 423
+    },
+    {
+      "epoch": 2.0809815950920245,
+      "grad_norm": 0.007294974289834499,
+      "learning_rate": 4.096039603960396e-06,
+      "loss": 0.0013,
+      "step": 424
+    },
+    {
+      "epoch": 2.085889570552147,
+      "grad_norm": 0.09458526223897934,
+      "learning_rate": 4.089108910891089e-06,
+      "loss": 0.0018,
+      "step": 425
+    },
+    {
+      "epoch": 2.09079754601227,
+      "grad_norm": 0.02894427254796028,
+      "learning_rate": 4.082178217821782e-06,
+      "loss": 0.0016,
+      "step": 426
+    },
+    {
+      "epoch": 2.0957055214723925,
+      "grad_norm": 0.030396826565265656,
+      "learning_rate": 4.075247524752475e-06,
+      "loss": 0.0018,
+      "step": 427
+    },
+    {
+      "epoch": 2.100613496932515,
+      "grad_norm": 0.11630258709192276,
+      "learning_rate": 4.068316831683168e-06,
+      "loss": 0.0081,
+      "step": 428
+    },
+    {
+      "epoch": 2.105521472392638,
+      "grad_norm": 0.010553350672125816,
+      "learning_rate": 4.061386138613861e-06,
+      "loss": 0.0008,
+      "step": 429
+    },
+    {
+      "epoch": 2.1104294478527605,
+      "grad_norm": 0.04130052775144577,
+      "learning_rate": 4.054455445544554e-06,
+      "loss": 0.0026,
+      "step": 430
+    },
+    {
+      "epoch": 2.1153374233128837,
+      "grad_norm": 0.10501858592033386,
+      "learning_rate": 4.047524752475247e-06,
+      "loss": 0.0396,
+      "step": 431
+    },
+    {
+      "epoch": 2.1202453987730063,
+      "grad_norm": 0.13223466277122498,
+      "learning_rate": 4.04059405940594e-06,
+      "loss": 0.0111,
+      "step": 432
+    },
+    {
+      "epoch": 2.125153374233129,
+      "grad_norm": 0.015712805092334747,
+      "learning_rate": 4.033663366336633e-06,
+      "loss": 0.0012,
+      "step": 433
+    },
+    {
+      "epoch": 2.1300613496932517,
+      "grad_norm": 0.023326637223362923,
+      "learning_rate": 4.026732673267326e-06,
+      "loss": 0.0017,
+      "step": 434
+    },
+    {
+      "epoch": 2.1349693251533743,
+      "grad_norm": 0.03263499587774277,
+      "learning_rate": 4.019801980198019e-06,
+      "loss": 0.002,
+      "step": 435
+    },
+    {
+      "epoch": 2.139877300613497,
+      "grad_norm": 0.018768969923257828,
+      "learning_rate": 4.0128712871287124e-06,
+      "loss": 0.0015,
+      "step": 436
+    },
+    {
+      "epoch": 2.1447852760736197,
+      "grad_norm": 0.4300386905670166,
+      "learning_rate": 4.0059405940594055e-06,
+      "loss": 0.064,
+      "step": 437
+    },
+    {
+      "epoch": 2.1496932515337424,
+      "grad_norm": 0.01180424727499485,
+      "learning_rate": 3.999009900990099e-06,
+      "loss": 0.0018,
+      "step": 438
+    },
+    {
+      "epoch": 2.154601226993865,
+      "grad_norm": 0.007693049497902393,
+      "learning_rate": 3.992079207920792e-06,
+      "loss": 0.0007,
+      "step": 439
+    },
+    {
+      "epoch": 2.1595092024539877,
+      "grad_norm": 0.04772064834833145,
+      "learning_rate": 3.985148514851485e-06,
+      "loss": 0.0031,
+      "step": 440
+    },
+    {
+      "epoch": 2.1644171779141104,
+      "grad_norm": 0.017260396853089333,
+      "learning_rate": 3.978217821782178e-06,
+      "loss": 0.001,
+      "step": 441
+    },
+    {
+      "epoch": 2.169325153374233,
+      "grad_norm": 0.08507797867059708,
+      "learning_rate": 3.971287128712871e-06,
+      "loss": 0.0063,
+      "step": 442
+    },
+    {
+      "epoch": 2.1742331288343557,
+      "grad_norm": 0.02232922799885273,
+      "learning_rate": 3.964356435643564e-06,
+      "loss": 0.002,
+      "step": 443
+    },
+    {
+      "epoch": 2.1791411042944784,
+      "grad_norm": 0.06088268384337425,
+      "learning_rate": 3.957425742574257e-06,
+      "loss": 0.0058,
+      "step": 444
+    },
+    {
+      "epoch": 2.184049079754601,
+      "grad_norm": 0.15273047983646393,
+      "learning_rate": 3.95049504950495e-06,
+      "loss": 0.0035,
+      "step": 445
+    },
+    {
+      "epoch": 2.1889570552147237,
+      "grad_norm": 0.014187682420015335,
+      "learning_rate": 3.943564356435643e-06,
+      "loss": 0.0014,
+      "step": 446
+    },
+    {
+      "epoch": 2.1938650306748464,
+      "grad_norm": 0.1353190392255783,
+      "learning_rate": 3.936633663366337e-06,
+      "loss": 0.003,
+      "step": 447
+    },
+    {
+      "epoch": 2.1987730061349695,
+      "grad_norm": 0.1862998902797699,
+      "learning_rate": 3.92970297029703e-06,
+      "loss": 0.0232,
+      "step": 448
+    },
+    {
+      "epoch": 2.203680981595092,
+      "grad_norm": 0.06980832666158676,
+      "learning_rate": 3.922772277227723e-06,
+      "loss": 0.0018,
+      "step": 449
+    },
+    {
+      "epoch": 2.208588957055215,
+      "grad_norm": 0.012548700906336308,
+      "learning_rate": 3.915841584158416e-06,
+      "loss": 0.0009,
+      "step": 450
+    },
+    {
+      "epoch": 2.2134969325153375,
+      "grad_norm": 0.0944492369890213,
+      "learning_rate": 3.908910891089109e-06,
+      "loss": 0.0071,
+      "step": 451
+    },
+    {
+      "epoch": 2.21840490797546,
+      "grad_norm": 0.07349948585033417,
+      "learning_rate": 3.901980198019802e-06,
+      "loss": 0.0073,
+      "step": 452
+    },
+    {
+      "epoch": 2.223312883435583,
+      "grad_norm": 0.016509253531694412,
+      "learning_rate": 3.8950495049504955e-06,
+      "loss": 0.0012,
+      "step": 453
+    },
+    {
+      "epoch": 2.2282208588957055,
+      "grad_norm": 0.0530993714928627,
+      "learning_rate": 3.8881188118811885e-06,
+      "loss": 0.0025,
+      "step": 454
+    },
+    {
+      "epoch": 2.233128834355828,
+      "grad_norm": 0.014129874296486378,
+      "learning_rate": 3.881188118811882e-06,
+      "loss": 0.002,
+      "step": 455
+    },
+    {
+      "epoch": 2.238036809815951,
+      "grad_norm": 0.16294129192829132,
+      "learning_rate": 3.874257425742575e-06,
+      "loss": 0.0049,
+      "step": 456
+    },
+    {
+      "epoch": 2.2429447852760735,
+      "grad_norm": 0.03273903578519821,
+      "learning_rate": 3.867326732673268e-06,
+      "loss": 0.0019,
+      "step": 457
+    },
+    {
+      "epoch": 2.247852760736196,
+      "grad_norm": 0.04015541449189186,
+      "learning_rate": 3.860396039603961e-06,
+      "loss": 0.0021,
+      "step": 458
+    },
+    {
+      "epoch": 2.252760736196319,
+      "grad_norm": 0.02188378944993019,
+      "learning_rate": 3.853465346534654e-06,
+      "loss": 0.001,
+      "step": 459
+    },
+    {
+      "epoch": 2.2576687116564416,
+      "grad_norm": 0.02690410614013672,
+      "learning_rate": 3.846534653465347e-06,
+      "loss": 0.0013,
+      "step": 460
+    },
+    {
+      "epoch": 2.2625766871165642,
+      "grad_norm": 0.08112508058547974,
+      "learning_rate": 3.83960396039604e-06,
+      "loss": 0.0027,
+      "step": 461
+    },
+    {
+      "epoch": 2.2674846625766873,
+      "grad_norm": 0.015935292467474937,
+      "learning_rate": 3.832673267326733e-06,
+      "loss": 0.0024,
+      "step": 462
+    },
+    {
+      "epoch": 2.27239263803681,
+      "grad_norm": 0.26772162318229675,
+      "learning_rate": 3.825742574257426e-06,
+      "loss": 0.0118,
+      "step": 463
+    },
+    {
+      "epoch": 2.2773006134969327,
+      "grad_norm": 0.035102855414152145,
+      "learning_rate": 3.818811881188119e-06,
+      "loss": 0.001,
+      "step": 464
+    },
+    {
+      "epoch": 2.2822085889570554,
+      "grad_norm": 0.03706807270646095,
+      "learning_rate": 3.8118811881188123e-06,
+      "loss": 0.0019,
+      "step": 465
+    },
+    {
+      "epoch": 2.287116564417178,
+      "grad_norm": 0.012720320373773575,
+      "learning_rate": 3.8049504950495054e-06,
+      "loss": 0.0012,
+      "step": 466
+    },
+    {
+      "epoch": 2.2920245398773007,
+      "grad_norm": 0.07940246909856796,
+      "learning_rate": 3.7980198019801984e-06,
+      "loss": 0.0045,
+      "step": 467
+    },
+    {
+      "epoch": 2.2969325153374234,
+      "grad_norm": 0.017415238544344902,
+      "learning_rate": 3.7910891089108915e-06,
+      "loss": 0.0012,
+      "step": 468
+    },
+    {
+      "epoch": 2.301840490797546,
+      "grad_norm": 0.1183788999915123,
+      "learning_rate": 3.7841584158415846e-06,
+      "loss": 0.0076,
+      "step": 469
+    },
+    {
+      "epoch": 2.3067484662576687,
+      "grad_norm": 0.01171356625854969,
+      "learning_rate": 3.7772277227722776e-06,
+      "loss": 0.001,
+      "step": 470
+    },
+    {
+      "epoch": 2.3116564417177914,
+      "grad_norm": 0.010136552155017853,
+      "learning_rate": 3.7702970297029703e-06,
+      "loss": 0.0009,
+      "step": 471
+    },
+    {
+      "epoch": 2.316564417177914,
+      "grad_norm": 0.028086962178349495,
+      "learning_rate": 3.7633663366336633e-06,
+      "loss": 0.0011,
+      "step": 472
+    },
+    {
+      "epoch": 2.3214723926380367,
+      "grad_norm": 0.00958373211324215,
+      "learning_rate": 3.7564356435643564e-06,
+      "loss": 0.0014,
+      "step": 473
+    },
+    {
+      "epoch": 2.3263803680981594,
+      "grad_norm": 0.034549593925476074,
+      "learning_rate": 3.7495049504950495e-06,
+      "loss": 0.0016,
+      "step": 474
+    },
+    {
+      "epoch": 2.331288343558282,
+      "grad_norm": 0.02567419782280922,
+      "learning_rate": 3.7425742574257425e-06,
+      "loss": 0.0017,
+      "step": 475
+    },
+    {
+      "epoch": 2.3361963190184047,
+      "grad_norm": 0.0217987522482872,
+      "learning_rate": 3.7356435643564356e-06,
+      "loss": 0.0012,
+      "step": 476
+    },
+    {
+      "epoch": 2.3411042944785274,
+      "grad_norm": 0.00880768895149231,
+      "learning_rate": 3.7287128712871287e-06,
+      "loss": 0.0012,
+      "step": 477
+    },
+    {
+      "epoch": 2.34601226993865,
+      "grad_norm": 0.0350213460624218,
+      "learning_rate": 3.7217821782178218e-06,
+      "loss": 0.0038,
+      "step": 478
+    },
+    {
+      "epoch": 2.3509202453987728,
+      "grad_norm": 0.009084700606763363,
+      "learning_rate": 3.714851485148515e-06,
+      "loss": 0.0018,
+      "step": 479
+    },
+    {
+      "epoch": 2.355828220858896,
+      "grad_norm": 0.056382663547992706,
+      "learning_rate": 3.707920792079208e-06,
+      "loss": 0.0018,
+      "step": 480
+    },
+    {
+      "epoch": 2.3607361963190185,
+      "grad_norm": 0.03614302724599838,
+      "learning_rate": 3.700990099009901e-06,
+      "loss": 0.0025,
+      "step": 481
+    },
+    {
+      "epoch": 2.365644171779141,
+      "grad_norm": 0.03676525875926018,
+      "learning_rate": 3.694059405940594e-06,
+      "loss": 0.001,
+      "step": 482
+    },
+    {
+      "epoch": 2.370552147239264,
+      "grad_norm": 0.1925876885652542,
+      "learning_rate": 3.687128712871287e-06,
+      "loss": 0.0141,
+      "step": 483
+    },
+    {
+      "epoch": 2.3754601226993866,
+      "grad_norm": 0.013960366137325764,
+      "learning_rate": 3.68019801980198e-06,
+      "loss": 0.0012,
+      "step": 484
+    },
+    {
+      "epoch": 2.3803680981595092,
+      "grad_norm": 0.03343435749411583,
+      "learning_rate": 3.6732673267326732e-06,
+      "loss": 0.0014,
+      "step": 485
+    },
+    {
+      "epoch": 2.385276073619632,
+      "grad_norm": 0.022218871861696243,
+      "learning_rate": 3.6663366336633663e-06,
+      "loss": 0.0026,
+      "step": 486
+    },
+    {
+      "epoch": 2.3901840490797546,
+      "grad_norm": 0.015421630814671516,
+      "learning_rate": 3.6594059405940594e-06,
+      "loss": 0.001,
+      "step": 487
+    },
+    {
+      "epoch": 2.3950920245398772,
+      "grad_norm": 0.011735438369214535,
+      "learning_rate": 3.6524752475247524e-06,
+      "loss": 0.0011,
+      "step": 488
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.06969886273145676,
+      "learning_rate": 3.6455445544554455e-06,
+      "loss": 0.0035,
+      "step": 489
+    },
+    {
+      "epoch": 2.4049079754601226,
+      "grad_norm": 0.009630633518099785,
+      "learning_rate": 3.6386138613861386e-06,
+      "loss": 0.0012,
+      "step": 490
+    },
+    {
+      "epoch": 2.4098159509202453,
+      "grad_norm": 0.18442873656749725,
+      "learning_rate": 3.6316831683168316e-06,
+      "loss": 0.0034,
+      "step": 491
+    },
+    {
+      "epoch": 2.414723926380368,
+      "grad_norm": 0.0310515183955431,
+      "learning_rate": 3.6247524752475247e-06,
+      "loss": 0.002,
+      "step": 492
+    },
+    {
+      "epoch": 2.419631901840491,
+      "grad_norm": 0.107456274330616,
+      "learning_rate": 3.6178217821782178e-06,
+      "loss": 0.0036,
+      "step": 493
+    },
+    {
+      "epoch": 2.4245398773006137,
+      "grad_norm": 0.018645675852894783,
+      "learning_rate": 3.610891089108911e-06,
+      "loss": 0.0018,
+      "step": 494
+    },
+    {
+      "epoch": 2.4294478527607364,
+      "grad_norm": 0.03003678098320961,
+      "learning_rate": 3.603960396039604e-06,
+      "loss": 0.0026,
+      "step": 495
+    },
+    {
+      "epoch": 2.434355828220859,
+      "grad_norm": 0.00237216055393219,
+      "learning_rate": 3.597029702970297e-06,
+      "loss": 0.0014,
+      "step": 496
+    },
+    {
+      "epoch": 2.4392638036809817,
+      "grad_norm": 0.11389046162366867,
+      "learning_rate": 3.59009900990099e-06,
+      "loss": 0.0036,
+      "step": 497
+    },
+    {
+      "epoch": 2.4441717791411044,
+      "grad_norm": 0.16872182488441467,
+      "learning_rate": 3.583168316831683e-06,
+      "loss": 0.0127,
+      "step": 498
+    },
+    {
+      "epoch": 2.449079754601227,
+      "grad_norm": 0.011766748502850533,
+      "learning_rate": 3.576237623762376e-06,
+      "loss": 0.0006,
+      "step": 499
+    },
+    {
+      "epoch": 2.4539877300613497,
+      "grad_norm": 0.014331258833408356,
+      "learning_rate": 3.5693069306930693e-06,
+      "loss": 0.0014,
+      "step": 500
+    },
+    {
+      "epoch": 2.4588957055214724,
+      "grad_norm": 0.009422739036381245,
+      "learning_rate": 3.5623762376237623e-06,
+      "loss": 0.0012,
+      "step": 501
+    },
+    {
+      "epoch": 2.463803680981595,
+      "grad_norm": 0.04546496644616127,
+      "learning_rate": 3.5554455445544554e-06,
+      "loss": 0.0067,
+      "step": 502
+    },
+    {
+      "epoch": 2.4687116564417177,
+      "grad_norm": 0.026389574632048607,
+      "learning_rate": 3.5485148514851485e-06,
+      "loss": 0.0016,
+      "step": 503
+    },
+    {
+      "epoch": 2.4736196319018404,
+      "grad_norm": 0.025073140859603882,
+      "learning_rate": 3.5415841584158415e-06,
+      "loss": 0.0011,
+      "step": 504
+    },
+    {
+      "epoch": 2.478527607361963,
+      "grad_norm": 0.032442186027765274,
+      "learning_rate": 3.5346534653465346e-06,
+      "loss": 0.0018,
+      "step": 505
+    },
+    {
+      "epoch": 2.4834355828220858,
+      "grad_norm": 0.07158058136701584,
+      "learning_rate": 3.5277227722772277e-06,
+      "loss": 0.003,
+      "step": 506
+    },
+    {
+      "epoch": 2.4883435582822084,
+      "grad_norm": 0.005822812672704458,
+      "learning_rate": 3.5207920792079207e-06,
+      "loss": 0.0009,
+      "step": 507
+    },
+    {
+      "epoch": 2.493251533742331,
+      "grad_norm": 0.0328555554151535,
+      "learning_rate": 3.513861386138614e-06,
+      "loss": 0.0013,
+      "step": 508
+    },
+    {
+      "epoch": 2.4981595092024538,
+      "grad_norm": 0.010760471224784851,
+      "learning_rate": 3.506930693069307e-06,
+      "loss": 0.001,
+      "step": 509
+    },
+    {
+      "epoch": 2.5030674846625764,
+      "grad_norm": 0.02453591674566269,
+      "learning_rate": 3.5e-06,
+      "loss": 0.0015,
+      "step": 510
+    },
+    {
+      "epoch": 2.507975460122699,
+      "grad_norm": 0.025182075798511505,
+      "learning_rate": 3.493069306930693e-06,
+      "loss": 0.0015,
+      "step": 511
+    },
+    {
+      "epoch": 2.5128834355828222,
+      "grad_norm": 0.25042542815208435,
+      "learning_rate": 3.486138613861386e-06,
+      "loss": 0.009,
+      "step": 512
+    },
+    {
+      "epoch": 2.517791411042945,
+      "grad_norm": 0.03413018584251404,
+      "learning_rate": 3.479207920792079e-06,
+      "loss": 0.0018,
+      "step": 513
+    },
+    {
+      "epoch": 2.5226993865030676,
+      "grad_norm": 0.16198301315307617,
+      "learning_rate": 3.4722772277227722e-06,
+      "loss": 0.0134,
+      "step": 514
+    },
+    {
+      "epoch": 2.5276073619631902,
+      "grad_norm": 0.17140725255012512,
+      "learning_rate": 3.4653465346534653e-06,
+      "loss": 0.0338,
+      "step": 515
+    },
+    {
+      "epoch": 2.532515337423313,
+      "grad_norm": 0.06337208300828934,
+      "learning_rate": 3.4584158415841584e-06,
+      "loss": 0.0042,
+      "step": 516
+    },
+    {
+      "epoch": 2.5374233128834356,
+      "grad_norm": 0.00469438498839736,
+      "learning_rate": 3.4514851485148514e-06,
+      "loss": 0.0005,
+      "step": 517
+    },
+    {
+      "epoch": 2.5423312883435583,
+      "grad_norm": 0.053614478558301926,
+      "learning_rate": 3.4445544554455445e-06,
+      "loss": 0.0063,
+      "step": 518
+    },
+    {
+      "epoch": 2.547239263803681,
+      "grad_norm": 0.019361039623618126,
+      "learning_rate": 3.4376237623762376e-06,
+      "loss": 0.0011,
+      "step": 519
+    },
+    {
+      "epoch": 2.5521472392638036,
+      "grad_norm": 0.017689630389213562,
+      "learning_rate": 3.4306930693069306e-06,
+      "loss": 0.0011,
+      "step": 520
+    },
+    {
+      "epoch": 2.5570552147239263,
+      "grad_norm": 0.03599102422595024,
+      "learning_rate": 3.4237623762376237e-06,
+      "loss": 0.0021,
+      "step": 521
+    },
+    {
+      "epoch": 2.561963190184049,
+      "grad_norm": 0.6623808145523071,
+      "learning_rate": 3.4168316831683168e-06,
+      "loss": 0.0231,
+      "step": 522
+    },
+    {
+      "epoch": 2.5668711656441716,
+      "grad_norm": 0.013597175478935242,
+      "learning_rate": 3.40990099009901e-06,
+      "loss": 0.0018,
+      "step": 523
+    },
+    {
+      "epoch": 2.5717791411042947,
+      "grad_norm": 0.07440067827701569,
+      "learning_rate": 3.402970297029703e-06,
+      "loss": 0.0043,
+      "step": 524
+    },
+    {
+      "epoch": 2.5766871165644174,
+      "grad_norm": 0.016382288187742233,
+      "learning_rate": 3.396039603960396e-06,
+      "loss": 0.0013,
+      "step": 525
+    },
+    {
+      "epoch": 2.58159509202454,
+      "grad_norm": 0.04037494957447052,
+      "learning_rate": 3.389108910891089e-06,
+      "loss": 0.0034,
+      "step": 526
+    },
+    {
+      "epoch": 2.5865030674846627,
+      "grad_norm": 0.08392734825611115,
+      "learning_rate": 3.382178217821782e-06,
+      "loss": 0.0022,
+      "step": 527
+    },
+    {
+      "epoch": 2.5914110429447854,
+      "grad_norm": 0.041856564581394196,
+      "learning_rate": 3.375247524752475e-06,
+      "loss": 0.0019,
+      "step": 528
+    },
+    {
+      "epoch": 2.596319018404908,
+      "grad_norm": 0.0511699840426445,
+      "learning_rate": 3.3683168316831683e-06,
+      "loss": 0.0016,
+      "step": 529
+    },
+    {
+      "epoch": 2.6012269938650308,
+      "grad_norm": 0.012307001277804375,
+      "learning_rate": 3.3613861386138613e-06,
+      "loss": 0.0011,
+      "step": 530
+    },
+    {
+      "epoch": 2.6061349693251534,
+      "grad_norm": 0.033212386071681976,
+      "learning_rate": 3.3544554455445544e-06,
+      "loss": 0.0018,
+      "step": 531
+    },
+    {
+      "epoch": 2.611042944785276,
+      "grad_norm": 0.033674679696559906,
+      "learning_rate": 3.3475247524752475e-06,
+      "loss": 0.0009,
+      "step": 532
+    },
+    {
+      "epoch": 2.6159509202453988,
+      "grad_norm": 0.022887440398335457,
+      "learning_rate": 3.3405940594059405e-06,
+      "loss": 0.0009,
+      "step": 533
+    },
+    {
+      "epoch": 2.6208588957055214,
+      "grad_norm": 0.08867110311985016,
+      "learning_rate": 3.3336633663366336e-06,
+      "loss": 0.0033,
+      "step": 534
+    },
+    {
+      "epoch": 2.625766871165644,
+      "grad_norm": 0.015654679387807846,
+      "learning_rate": 3.3267326732673267e-06,
+      "loss": 0.0011,
+      "step": 535
+    },
+    {
+      "epoch": 2.630674846625767,
+      "grad_norm": 0.07898303866386414,
+      "learning_rate": 3.3198019801980197e-06,
+      "loss": 0.0029,
+      "step": 536
+    },
+    {
+      "epoch": 2.6355828220858895,
+      "grad_norm": 0.1686147004365921,
+      "learning_rate": 3.312871287128713e-06,
+      "loss": 0.0115,
+      "step": 537
+    },
+    {
+      "epoch": 2.640490797546012,
+      "grad_norm": 0.03158680722117424,
+      "learning_rate": 3.305940594059406e-06,
+      "loss": 0.0017,
+      "step": 538
+    },
+    {
+      "epoch": 2.645398773006135,
+      "grad_norm": 0.11354830116033554,
+      "learning_rate": 3.299009900990099e-06,
+      "loss": 0.0091,
+      "step": 539
+    },
+    {
+      "epoch": 2.6503067484662575,
+      "grad_norm": 0.08498039096593857,
+      "learning_rate": 3.292079207920792e-06,
+      "loss": 0.0038,
+      "step": 540
+    },
+    {
+      "epoch": 2.65521472392638,
+      "grad_norm": 0.08356107771396637,
+      "learning_rate": 3.285148514851485e-06,
+      "loss": 0.0019,
+      "step": 541
+    },
+    {
+      "epoch": 2.660122699386503,
+      "grad_norm": 0.24042771756649017,
+      "learning_rate": 3.2782178217821786e-06,
+      "loss": 0.0071,
+      "step": 542
+    },
+    {
+      "epoch": 2.665030674846626,
+      "grad_norm": 0.034460801631212234,
+      "learning_rate": 3.2712871287128716e-06,
+      "loss": 0.0017,
+      "step": 543
+    },
+    {
+      "epoch": 2.6699386503067486,
+      "grad_norm": 0.09137725085020065,
+      "learning_rate": 3.2643564356435643e-06,
+      "loss": 0.0022,
+      "step": 544
+    },
+    {
+      "epoch": 2.6748466257668713,
+      "grad_norm": 0.04045404493808746,
+      "learning_rate": 3.2574257425742573e-06,
+      "loss": 0.0016,
+      "step": 545
+    },
+    {
+      "epoch": 2.679754601226994,
+      "grad_norm": 0.24448804557323456,
+      "learning_rate": 3.2504950495049504e-06,
+      "loss": 0.0228,
+      "step": 546
+    },
+    {
+      "epoch": 2.6846625766871166,
+      "grad_norm": 0.02107204869389534,
+      "learning_rate": 3.2435643564356435e-06,
+      "loss": 0.0022,
+      "step": 547
+    },
+    {
+      "epoch": 2.6895705521472393,
+      "grad_norm": 0.02561573125422001,
+      "learning_rate": 3.2366336633663366e-06,
+      "loss": 0.0014,
+      "step": 548
+    },
+    {
+      "epoch": 2.694478527607362,
+      "grad_norm": 0.02062409184873104,
+      "learning_rate": 3.2297029702970296e-06,
+      "loss": 0.0009,
+      "step": 549
+    },
+    {
+      "epoch": 2.6993865030674846,
+      "grad_norm": 0.053445011377334595,
+      "learning_rate": 3.2227722772277227e-06,
+      "loss": 0.0025,
+      "step": 550
+    },
+    {
+      "epoch": 2.7042944785276073,
+      "grad_norm": 0.0313737653195858,
+      "learning_rate": 3.2158415841584158e-06,
+      "loss": 0.0021,
+      "step": 551
+    },
+    {
+      "epoch": 2.70920245398773,
+      "grad_norm": 0.04768161103129387,
+      "learning_rate": 3.208910891089109e-06,
+      "loss": 0.006,
+      "step": 552
+    },
+    {
+      "epoch": 2.7141104294478526,
+      "grad_norm": 0.1395123153924942,
+      "learning_rate": 3.201980198019802e-06,
+      "loss": 0.0095,
+      "step": 553
+    },
+    {
+      "epoch": 2.7190184049079753,
+      "grad_norm": 0.022861650213599205,
+      "learning_rate": 3.195049504950495e-06,
+      "loss": 0.0014,
+      "step": 554
+    },
+    {
+      "epoch": 2.7239263803680984,
+      "grad_norm": 0.08337707072496414,
+      "learning_rate": 3.188118811881188e-06,
+      "loss": 0.0053,
+      "step": 555
+    },
+    {
+      "epoch": 2.728834355828221,
+      "grad_norm": 0.1470242589712143,
+      "learning_rate": 3.181188118811881e-06,
+      "loss": 0.0089,
+      "step": 556
+    },
+    {
+      "epoch": 2.7337423312883438,
+      "grad_norm": 0.1716291755437851,
+      "learning_rate": 3.174257425742574e-06,
+      "loss": 0.014,
+      "step": 557
+    },
+    {
+      "epoch": 2.7386503067484664,
+      "grad_norm": 0.015148639678955078,
+      "learning_rate": 3.1673267326732672e-06,
+      "loss": 0.001,
+      "step": 558
+    },
+    {
+      "epoch": 2.743558282208589,
+      "grad_norm": 0.019988784566521645,
+      "learning_rate": 3.1603960396039603e-06,
+      "loss": 0.0015,
+      "step": 559
+    },
+    {
+      "epoch": 2.7484662576687118,
+      "grad_norm": 0.2989899814128876,
+      "learning_rate": 3.1534653465346534e-06,
+      "loss": 0.0092,
+      "step": 560
+    },
+    {
+      "epoch": 2.7533742331288344,
+      "grad_norm": 0.1491839736700058,
+      "learning_rate": 3.1465346534653464e-06,
+      "loss": 0.0046,
+      "step": 561
+    },
+    {
+      "epoch": 2.758282208588957,
+      "grad_norm": 0.02539440244436264,
+      "learning_rate": 3.1396039603960395e-06,
+      "loss": 0.0014,
+      "step": 562
+    },
+    {
+      "epoch": 2.76319018404908,
+      "grad_norm": 0.31437137722969055,
+      "learning_rate": 3.1326732673267326e-06,
+      "loss": 0.0109,
+      "step": 563
+    },
+    {
+      "epoch": 2.7680981595092025,
+      "grad_norm": 0.013294244185090065,
+      "learning_rate": 3.1257425742574257e-06,
+      "loss": 0.0016,
+      "step": 564
+    },
+    {
+      "epoch": 2.773006134969325,
+      "grad_norm": 0.025412971153855324,
+      "learning_rate": 3.1188118811881187e-06,
+      "loss": 0.002,
+      "step": 565
+    },
+    {
+      "epoch": 2.777914110429448,
+      "grad_norm": 0.05206981673836708,
+      "learning_rate": 3.1118811881188118e-06,
+      "loss": 0.0017,
+      "step": 566
+    },
+    {
+      "epoch": 2.7828220858895705,
+      "grad_norm": 0.0227506086230278,
+      "learning_rate": 3.104950495049505e-06,
+      "loss": 0.001,
+      "step": 567
+    },
+    {
+      "epoch": 2.787730061349693,
+      "grad_norm": 0.019159123301506042,
+      "learning_rate": 3.098019801980198e-06,
+      "loss": 0.0012,
+      "step": 568
+    },
+    {
+      "epoch": 2.792638036809816,
+      "grad_norm": 0.035373397171497345,
+      "learning_rate": 3.091089108910891e-06,
+      "loss": 0.0011,
+      "step": 569
+    },
+    {
+      "epoch": 2.7975460122699385,
+      "grad_norm": 0.06474697589874268,
+      "learning_rate": 3.084158415841584e-06,
+      "loss": 0.0017,
+      "step": 570
+    },
+    {
+      "epoch": 2.802453987730061,
+      "grad_norm": 0.01330810971558094,
+      "learning_rate": 3.077227722772277e-06,
+      "loss": 0.0006,
+      "step": 571
+    },
+    {
+      "epoch": 2.807361963190184,
+      "grad_norm": 0.018572993576526642,
+      "learning_rate": 3.07029702970297e-06,
+      "loss": 0.0012,
+      "step": 572
+    },
+    {
+      "epoch": 2.8122699386503065,
+      "grad_norm": 0.04764172062277794,
+      "learning_rate": 3.0633663366336633e-06,
+      "loss": 0.0027,
+      "step": 573
+    },
+    {
+      "epoch": 2.817177914110429,
+      "grad_norm": 0.05075710266828537,
+      "learning_rate": 3.0564356435643563e-06,
+      "loss": 0.0018,
+      "step": 574
+    },
+    {
+      "epoch": 2.8220858895705523,
+      "grad_norm": 0.0751049742102623,
+      "learning_rate": 3.0495049504950494e-06,
+      "loss": 0.0023,
+      "step": 575
+    },
+    {
+      "epoch": 2.826993865030675,
+      "grad_norm": 0.21285609900951385,
+      "learning_rate": 3.0425742574257425e-06,
+      "loss": 0.0089,
+      "step": 576
+    },
+    {
+      "epoch": 2.8319018404907976,
+      "grad_norm": 0.048560746014118195,
+      "learning_rate": 3.0356435643564355e-06,
+      "loss": 0.0029,
+      "step": 577
+    },
+    {
+      "epoch": 2.8368098159509203,
+      "grad_norm": 0.09979841113090515,
+      "learning_rate": 3.0287128712871286e-06,
+      "loss": 0.0035,
+      "step": 578
+    },
+    {
+      "epoch": 2.841717791411043,
+      "grad_norm": 0.012869380414485931,
+      "learning_rate": 3.0217821782178217e-06,
+      "loss": 0.0011,
+      "step": 579
+    },
+    {
+      "epoch": 2.8466257668711656,
+      "grad_norm": 0.04724825918674469,
+      "learning_rate": 3.0148514851485147e-06,
+      "loss": 0.0021,
+      "step": 580
+    },
+    {
+      "epoch": 2.8515337423312883,
+      "grad_norm": 0.10769210010766983,
+      "learning_rate": 3.007920792079208e-06,
+      "loss": 0.0097,
+      "step": 581
+    },
+    {
+      "epoch": 2.856441717791411,
+      "grad_norm": 0.00834321416914463,
+      "learning_rate": 3.000990099009901e-06,
+      "loss": 0.0008,
+      "step": 582
+    },
+    {
+      "epoch": 2.8613496932515337,
+      "grad_norm": 0.04464031755924225,
+      "learning_rate": 2.994059405940594e-06,
+      "loss": 0.0026,
+      "step": 583
+    },
+    {
+      "epoch": 2.8662576687116563,
+      "grad_norm": 0.01902065984904766,
+      "learning_rate": 2.987128712871287e-06,
+      "loss": 0.0014,
+      "step": 584
+    },
+    {
+      "epoch": 2.871165644171779,
+      "grad_norm": 0.038237448781728745,
+      "learning_rate": 2.98019801980198e-06,
+      "loss": 0.0015,
+      "step": 585
+    },
+    {
+      "epoch": 2.876073619631902,
+      "grad_norm": 0.013939250260591507,
+      "learning_rate": 2.973267326732673e-06,
+      "loss": 0.001,
+      "step": 586
+    },
+    {
+      "epoch": 2.880981595092025,
+      "grad_norm": 0.03710507974028587,
+      "learning_rate": 2.9663366336633662e-06,
+      "loss": 0.0027,
+      "step": 587
+    },
+    {
+      "epoch": 2.8858895705521475,
+      "grad_norm": 0.04779844731092453,
+      "learning_rate": 2.9594059405940593e-06,
+      "loss": 0.0012,
+      "step": 588
+    },
+    {
+      "epoch": 2.89079754601227,
+      "grad_norm": 0.00711404625326395,
+      "learning_rate": 2.9524752475247524e-06,
+      "loss": 0.001,
+      "step": 589
+    },
+    {
+      "epoch": 2.895705521472393,
+      "grad_norm": 0.017887057736516,
+      "learning_rate": 2.9455445544554454e-06,
+      "loss": 0.0015,
+      "step": 590
+    },
+    {
+      "epoch": 2.9006134969325155,
+      "grad_norm": 0.01937202550470829,
+      "learning_rate": 2.9386138613861385e-06,
+      "loss": 0.0012,
+      "step": 591
+    },
+    {
+      "epoch": 2.905521472392638,
+      "grad_norm": 0.005528996232897043,
+      "learning_rate": 2.9316831683168316e-06,
+      "loss": 0.0011,
+      "step": 592
+    },
+    {
+      "epoch": 2.910429447852761,
+      "grad_norm": 0.025270938873291016,
+      "learning_rate": 2.9247524752475246e-06,
+      "loss": 0.0011,
+      "step": 593
+    },
+    {
+      "epoch": 2.9153374233128835,
+      "grad_norm": 0.022803837433457375,
+      "learning_rate": 2.9178217821782177e-06,
+      "loss": 0.0012,
+      "step": 594
+    },
+    {
+      "epoch": 2.920245398773006,
+      "grad_norm": 0.4284270405769348,
+      "learning_rate": 2.9108910891089108e-06,
+      "loss": 0.0206,
+      "step": 595
+    },
+    {
+      "epoch": 2.925153374233129,
+      "grad_norm": 0.03455421328544617,
+      "learning_rate": 2.903960396039604e-06,
+      "loss": 0.0016,
+      "step": 596
+    },
+    {
+      "epoch": 2.9300613496932515,
+      "grad_norm": 0.06015002727508545,
+      "learning_rate": 2.897029702970297e-06,
+      "loss": 0.0026,
+      "step": 597
+    },
+    {
+      "epoch": 2.934969325153374,
+      "grad_norm": 0.07220069319009781,
+      "learning_rate": 2.89009900990099e-06,
+      "loss": 0.0035,
+      "step": 598
+    },
+    {
+      "epoch": 2.939877300613497,
+      "grad_norm": 0.0011812745360657573,
+      "learning_rate": 2.883168316831683e-06,
+      "loss": 0.0003,
+      "step": 599
+    },
+    {
+      "epoch": 2.9447852760736195,
+      "grad_norm": 0.030697235837578773,
+      "learning_rate": 2.876237623762376e-06,
+      "loss": 0.0022,
+      "step": 600
+    },
+    {
+      "epoch": 2.949693251533742,
+      "grad_norm": 0.028459953144192696,
+      "learning_rate": 2.869306930693069e-06,
+      "loss": 0.0014,
+      "step": 601
+    },
+    {
+      "epoch": 2.954601226993865,
+      "grad_norm": 0.03279620781540871,
+      "learning_rate": 2.8623762376237623e-06,
+      "loss": 0.0019,
+      "step": 602
+    },
+    {
+      "epoch": 2.9595092024539875,
+      "grad_norm": 0.03629060834646225,
+      "learning_rate": 2.8554455445544553e-06,
+      "loss": 0.0023,
+      "step": 603
+    },
+    {
+      "epoch": 2.96441717791411,
+      "grad_norm": 0.008510543964803219,
+      "learning_rate": 2.8485148514851484e-06,
+      "loss": 0.0006,
+      "step": 604
+    },
+    {
+      "epoch": 2.969325153374233,
+      "grad_norm": 0.014275304973125458,
+      "learning_rate": 2.841584158415842e-06,
+      "loss": 0.001,
+      "step": 605
+    },
+    {
+      "epoch": 2.974233128834356,
+      "grad_norm": 0.06183558329939842,
+      "learning_rate": 2.834653465346535e-06,
+      "loss": 0.0021,
+      "step": 606
+    },
+    {
+      "epoch": 2.9791411042944786,
+      "grad_norm": 0.07287970185279846,
+      "learning_rate": 2.827722772277228e-06,
+      "loss": 0.0029,
+      "step": 607
+    },
+    {
+      "epoch": 2.9840490797546013,
+      "grad_norm": 0.02359975501894951,
+      "learning_rate": 2.820792079207921e-06,
+      "loss": 0.0015,
+      "step": 608
+    },
+    {
+      "epoch": 2.988957055214724,
+      "grad_norm": 0.02087857946753502,
+      "learning_rate": 2.813861386138614e-06,
+      "loss": 0.0027,
+      "step": 609
+    },
+    {
+      "epoch": 2.9938650306748467,
+      "grad_norm": 0.036885183304548264,
+      "learning_rate": 2.8069306930693072e-06,
+      "loss": 0.0016,
+      "step": 610
+    },
+    {
+      "epoch": 2.9987730061349693,
+      "grad_norm": 0.08759594708681107,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.013,
+      "step": 611
+    },
+    {
+      "epoch": 3.003680981595092,
+      "grad_norm": 0.0203255582600832,
+      "learning_rate": 2.7930693069306934e-06,
+      "loss": 0.0046,
+      "step": 612
+    },
+    {
+      "epoch": 3.0085889570552147,
+      "grad_norm": 0.04549151286482811,
+      "learning_rate": 2.7861386138613864e-06,
+      "loss": 0.0032,
+      "step": 613
+    },
+    {
+      "epoch": 3.0134969325153373,
+      "grad_norm": 0.02024606615304947,
+      "learning_rate": 2.7792079207920795e-06,
+      "loss": 0.0014,
+      "step": 614
+    },
+    {
+      "epoch": 3.01840490797546,
+      "grad_norm": 0.03339584916830063,
+      "learning_rate": 2.7722772277227726e-06,
+      "loss": 0.0018,
+      "step": 615
+    },
+    {
+      "epoch": 3.0233128834355827,
+      "grad_norm": 0.02375354804098606,
+      "learning_rate": 2.7653465346534656e-06,
+      "loss": 0.0028,
+      "step": 616
+    },
+    {
+      "epoch": 3.0282208588957054,
+      "grad_norm": 0.021728744730353355,
+      "learning_rate": 2.7584158415841583e-06,
+      "loss": 0.0031,
+      "step": 617
+    },
+    {
+      "epoch": 3.033128834355828,
+      "grad_norm": 0.032941028475761414,
+      "learning_rate": 2.7514851485148514e-06,
+      "loss": 0.0018,
+      "step": 618
+    },
+    {
+      "epoch": 3.038036809815951,
+      "grad_norm": 0.020893137902021408,
+      "learning_rate": 2.7445544554455444e-06,
+      "loss": 0.0013,
+      "step": 619
+    },
+    {
+      "epoch": 3.042944785276074,
+      "grad_norm": 0.012155863456428051,
+      "learning_rate": 2.7376237623762375e-06,
+      "loss": 0.0013,
+      "step": 620
+    },
+    {
+      "epoch": 3.0478527607361965,
+      "grad_norm": 0.0235965047031641,
+      "learning_rate": 2.7306930693069306e-06,
+      "loss": 0.0019,
+      "step": 621
+    },
+    {
+      "epoch": 3.052760736196319,
+      "grad_norm": 0.03598389774560928,
+      "learning_rate": 2.7237623762376236e-06,
+      "loss": 0.002,
+      "step": 622
+    },
+    {
+      "epoch": 3.057668711656442,
+      "grad_norm": 0.033302754163742065,
+      "learning_rate": 2.7168316831683167e-06,
+      "loss": 0.0011,
+      "step": 623
+    },
+    {
+      "epoch": 3.0625766871165645,
+      "grad_norm": 0.011287711560726166,
+      "learning_rate": 2.7099009900990098e-06,
+      "loss": 0.0015,
+      "step": 624
+    },
+    {
+      "epoch": 3.067484662576687,
+      "grad_norm": 0.04231281206011772,
+      "learning_rate": 2.702970297029703e-06,
+      "loss": 0.0032,
+      "step": 625
+    },
+    {
+      "epoch": 3.07239263803681,
+      "grad_norm": 0.020790139213204384,
+      "learning_rate": 2.696039603960396e-06,
+      "loss": 0.001,
+      "step": 626
+    },
+    {
+      "epoch": 3.0773006134969325,
+      "grad_norm": 0.020422058179974556,
+      "learning_rate": 2.689108910891089e-06,
+      "loss": 0.0017,
+      "step": 627
+    },
+    {
+      "epoch": 3.082208588957055,
+      "grad_norm": 0.02271398715674877,
+      "learning_rate": 2.682178217821782e-06,
+      "loss": 0.0013,
+      "step": 628
+    },
+    {
+      "epoch": 3.087116564417178,
+      "grad_norm": 0.013147884979844093,
+      "learning_rate": 2.675247524752475e-06,
+      "loss": 0.0009,
+      "step": 629
+    },
+    {
+      "epoch": 3.0920245398773005,
+      "grad_norm": 0.023107754066586494,
+      "learning_rate": 2.668316831683168e-06,
+      "loss": 0.0021,
+      "step": 630
+    },
+    {
+      "epoch": 3.096932515337423,
+      "grad_norm": 0.027083350345492363,
+      "learning_rate": 2.6613861386138612e-06,
+      "loss": 0.0016,
+      "step": 631
+    },
+    {
+      "epoch": 3.101840490797546,
+      "grad_norm": 0.01890239305794239,
+      "learning_rate": 2.6544554455445543e-06,
+      "loss": 0.0021,
+      "step": 632
+    },
+    {
+      "epoch": 3.1067484662576685,
+      "grad_norm": 0.0020133615471422672,
+      "learning_rate": 2.6475247524752474e-06,
+      "loss": 0.0004,
+      "step": 633
+    },
+    {
+      "epoch": 3.111656441717791,
+      "grad_norm": 0.022084610536694527,
+      "learning_rate": 2.6405940594059405e-06,
+      "loss": 0.0015,
+      "step": 634
+    },
+    {
+      "epoch": 3.116564417177914,
+      "grad_norm": 0.008074000477790833,
+      "learning_rate": 2.6336633663366335e-06,
+      "loss": 0.0011,
+      "step": 635
+    },
+    {
+      "epoch": 3.121472392638037,
+      "grad_norm": 0.03898672014474869,
+      "learning_rate": 2.6267326732673266e-06,
+      "loss": 0.002,
+      "step": 636
+    },
+    {
+      "epoch": 3.1263803680981597,
+      "grad_norm": 0.00946191418915987,
+      "learning_rate": 2.6198019801980197e-06,
+      "loss": 0.001,
+      "step": 637
+    },
+    {
+      "epoch": 3.1312883435582823,
+      "grad_norm": 0.1219155415892601,
+      "learning_rate": 2.6128712871287127e-06,
+      "loss": 0.0079,
+      "step": 638
+    },
+    {
+      "epoch": 3.136196319018405,
+      "grad_norm": 0.025132469832897186,
+      "learning_rate": 2.605940594059406e-06,
+      "loss": 0.0015,
+      "step": 639
+    },
+    {
+      "epoch": 3.1411042944785277,
+      "grad_norm": 0.01082681119441986,
+      "learning_rate": 2.599009900990099e-06,
+      "loss": 0.0008,
+      "step": 640
+    },
+    {
+      "epoch": 3.1460122699386504,
+      "grad_norm": 0.14690011739730835,
+      "learning_rate": 2.592079207920792e-06,
+      "loss": 0.0117,
+      "step": 641
+    },
+    {
+      "epoch": 3.150920245398773,
+      "grad_norm": 0.010698092170059681,
+      "learning_rate": 2.585148514851485e-06,
+      "loss": 0.0013,
+      "step": 642
+    },
+    {
+      "epoch": 3.1558282208588957,
+      "grad_norm": 0.2589493691921234,
+      "learning_rate": 2.578217821782178e-06,
+      "loss": 0.0089,
+      "step": 643
+    },
+    {
+      "epoch": 3.1607361963190184,
+      "grad_norm": 0.01592063717544079,
+      "learning_rate": 2.571287128712871e-06,
+      "loss": 0.0011,
+      "step": 644
+    },
+    {
+      "epoch": 3.165644171779141,
+      "grad_norm": 0.02394460327923298,
+      "learning_rate": 2.564356435643564e-06,
+      "loss": 0.0014,
+      "step": 645
+    },
+    {
+      "epoch": 3.1705521472392637,
+      "grad_norm": 0.02034229226410389,
+      "learning_rate": 2.5574257425742573e-06,
+      "loss": 0.0023,
+      "step": 646
+    },
+    {
+      "epoch": 3.1754601226993864,
+      "grad_norm": 0.032998789101839066,
+      "learning_rate": 2.5504950495049503e-06,
+      "loss": 0.0021,
+      "step": 647
+    },
+    {
+      "epoch": 3.180368098159509,
+      "grad_norm": 0.013732580468058586,
+      "learning_rate": 2.5435643564356434e-06,
+      "loss": 0.0005,
+      "step": 648
+    },
+    {
+      "epoch": 3.1852760736196317,
+      "grad_norm": 0.046656981110572815,
+      "learning_rate": 2.5366336633663365e-06,
+      "loss": 0.0031,
+      "step": 649
+    },
+    {
+      "epoch": 3.190184049079755,
+      "grad_norm": 0.02960779331624508,
+      "learning_rate": 2.5297029702970295e-06,
+      "loss": 0.0013,
+      "step": 650
+    },
+    {
+      "epoch": 3.1950920245398775,
+      "grad_norm": 0.051251210272312164,
+      "learning_rate": 2.5227722772277226e-06,
+      "loss": 0.0025,
+      "step": 651
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.03321881592273712,
+      "learning_rate": 2.5158415841584157e-06,
+      "loss": 0.0047,
+      "step": 652
+    },
+    {
+      "epoch": 3.204907975460123,
+      "grad_norm": 0.005035923328250647,
+      "learning_rate": 2.5089108910891088e-06,
+      "loss": 0.0007,
+      "step": 653
+    },
+    {
+      "epoch": 3.2098159509202455,
+      "grad_norm": 0.07169622927904129,
+      "learning_rate": 2.501980198019802e-06,
+      "loss": 0.0026,
+      "step": 654
+    },
+    {
+      "epoch": 3.214723926380368,
+      "grad_norm": 0.017996247857809067,
+      "learning_rate": 2.495049504950495e-06,
+      "loss": 0.0012,
+      "step": 655
+    },
+    {
+      "epoch": 3.219631901840491,
+      "grad_norm": 0.025565218180418015,
+      "learning_rate": 2.488118811881188e-06,
+      "loss": 0.0013,
+      "step": 656
+    },
+    {
+      "epoch": 3.2245398773006135,
+      "grad_norm": 0.014627007767558098,
+      "learning_rate": 2.481188118811881e-06,
+      "loss": 0.0015,
+      "step": 657
+    },
+    {
+      "epoch": 3.229447852760736,
+      "grad_norm": 0.011715116910636425,
+      "learning_rate": 2.474257425742574e-06,
+      "loss": 0.001,
+      "step": 658
+    },
+    {
+      "epoch": 3.234355828220859,
+      "grad_norm": 0.01887853443622589,
+      "learning_rate": 2.467326732673267e-06,
+      "loss": 0.0013,
+      "step": 659
+    },
+    {
+      "epoch": 3.2392638036809815,
+      "grad_norm": 0.028289880603551865,
+      "learning_rate": 2.4603960396039602e-06,
+      "loss": 0.0024,
+      "step": 660
+    },
+    {
+      "epoch": 3.244171779141104,
+      "grad_norm": 0.15456917881965637,
+      "learning_rate": 2.4534653465346533e-06,
+      "loss": 0.021,
+      "step": 661
+    },
+    {
+      "epoch": 3.249079754601227,
+      "grad_norm": 0.012570716440677643,
+      "learning_rate": 2.4465346534653464e-06,
+      "loss": 0.0012,
+      "step": 662
+    },
+    {
+      "epoch": 3.2539877300613496,
+      "grad_norm": 0.006434513721615076,
+      "learning_rate": 2.4396039603960394e-06,
+      "loss": 0.0006,
+      "step": 663
+    },
+    {
+      "epoch": 3.2588957055214722,
+      "grad_norm": 0.003889314830303192,
+      "learning_rate": 2.4326732673267325e-06,
+      "loss": 0.0007,
+      "step": 664
+    },
+    {
+      "epoch": 3.263803680981595,
+      "grad_norm": 0.16952529549598694,
+      "learning_rate": 2.4257425742574256e-06,
+      "loss": 0.008,
+      "step": 665
+    },
+    {
+      "epoch": 3.2687116564417176,
+      "grad_norm": 0.02936733514070511,
+      "learning_rate": 2.4188118811881186e-06,
+      "loss": 0.0016,
+      "step": 666
+    },
+    {
+      "epoch": 3.2736196319018402,
+      "grad_norm": 0.10733254253864288,
+      "learning_rate": 2.4118811881188117e-06,
+      "loss": 0.0029,
+      "step": 667
+    },
+    {
+      "epoch": 3.2785276073619634,
+      "grad_norm": 0.02705569751560688,
+      "learning_rate": 2.404950495049505e-06,
+      "loss": 0.0017,
+      "step": 668
+    },
+    {
+      "epoch": 3.283435582822086,
+      "grad_norm": 0.024813305586576462,
+      "learning_rate": 2.3980198019801983e-06,
+      "loss": 0.0011,
+      "step": 669
+    },
+    {
+      "epoch": 3.2883435582822087,
+      "grad_norm": 0.11823663115501404,
+      "learning_rate": 2.3910891089108913e-06,
+      "loss": 0.0034,
+      "step": 670
+    },
+    {
+      "epoch": 3.2932515337423314,
+      "grad_norm": 0.15911535918712616,
+      "learning_rate": 2.3841584158415844e-06,
+      "loss": 0.0038,
+      "step": 671
+    },
+    {
+      "epoch": 3.298159509202454,
+      "grad_norm": 0.03651705011725426,
+      "learning_rate": 2.3772277227722775e-06,
+      "loss": 0.0013,
+      "step": 672
+    },
+    {
+      "epoch": 3.3030674846625767,
+      "grad_norm": 0.20837312936782837,
+      "learning_rate": 2.3702970297029705e-06,
+      "loss": 0.0076,
+      "step": 673
+    },
+    {
+      "epoch": 3.3079754601226994,
+      "grad_norm": 0.013823213055729866,
+      "learning_rate": 2.3633663366336636e-06,
+      "loss": 0.0013,
+      "step": 674
+    },
+    {
+      "epoch": 3.312883435582822,
+      "grad_norm": 0.0133537407964468,
+      "learning_rate": 2.3564356435643567e-06,
+      "loss": 0.0018,
+      "step": 675
+    },
+    {
+      "epoch": 3.3177914110429447,
+      "grad_norm": 0.01537949126213789,
+      "learning_rate": 2.3495049504950498e-06,
+      "loss": 0.0013,
+      "step": 676
+    },
+    {
+      "epoch": 3.3226993865030674,
+      "grad_norm": 0.03161802887916565,
+      "learning_rate": 2.342574257425743e-06,
+      "loss": 0.0013,
+      "step": 677
+    },
+    {
+      "epoch": 3.32760736196319,
+      "grad_norm": 0.01877240464091301,
+      "learning_rate": 2.335643564356436e-06,
+      "loss": 0.001,
+      "step": 678
+    },
+    {
+      "epoch": 3.3325153374233127,
+      "grad_norm": 0.20145629346370697,
+      "learning_rate": 2.328712871287129e-06,
+      "loss": 0.0099,
+      "step": 679
+    },
+    {
+      "epoch": 3.3374233128834354,
+      "grad_norm": 0.17651820182800293,
+      "learning_rate": 2.321782178217822e-06,
+      "loss": 0.0298,
+      "step": 680
+    },
+    {
+      "epoch": 3.3423312883435585,
+      "grad_norm": 0.031260013580322266,
+      "learning_rate": 2.314851485148515e-06,
+      "loss": 0.0017,
+      "step": 681
+    },
+    {
+      "epoch": 3.347239263803681,
+      "grad_norm": 0.049838535487651825,
+      "learning_rate": 2.307920792079208e-06,
+      "loss": 0.0016,
+      "step": 682
+    },
+    {
+      "epoch": 3.352147239263804,
+      "grad_norm": 0.01825704425573349,
+      "learning_rate": 2.3009900990099012e-06,
+      "loss": 0.0009,
+      "step": 683
+    },
+    {
+      "epoch": 3.3570552147239265,
+      "grad_norm": 0.2820407450199127,
+      "learning_rate": 2.2940594059405943e-06,
+      "loss": 0.0095,
+      "step": 684
+    },
+    {
+      "epoch": 3.361963190184049,
+      "grad_norm": 0.013412845320999622,
+      "learning_rate": 2.2871287128712874e-06,
+      "loss": 0.0016,
+      "step": 685
+    },
+    {
+      "epoch": 3.366871165644172,
+      "grad_norm": 0.043872177600860596,
+      "learning_rate": 2.2801980198019804e-06,
+      "loss": 0.0023,
+      "step": 686
+    },
+    {
+      "epoch": 3.3717791411042946,
+      "grad_norm": 0.024329353123903275,
+      "learning_rate": 2.2732673267326735e-06,
+      "loss": 0.001,
+      "step": 687
+    },
+    {
+      "epoch": 3.3766871165644172,
+      "grad_norm": 0.016059909015893936,
+      "learning_rate": 2.2663366336633666e-06,
+      "loss": 0.0012,
+      "step": 688
+    },
+    {
+      "epoch": 3.38159509202454,
+      "grad_norm": 0.014483323320746422,
+      "learning_rate": 2.2594059405940596e-06,
+      "loss": 0.0011,
+      "step": 689
+    },
+    {
+      "epoch": 3.3865030674846626,
+      "grad_norm": 0.05332216992974281,
+      "learning_rate": 2.2524752475247523e-06,
+      "loss": 0.0023,
+      "step": 690
+    },
+    {
+      "epoch": 3.3914110429447852,
+      "grad_norm": 0.005560703109949827,
+      "learning_rate": 2.2455445544554454e-06,
+      "loss": 0.0006,
+      "step": 691
+    },
+    {
+      "epoch": 3.396319018404908,
+      "grad_norm": 0.029570411890745163,
+      "learning_rate": 2.2386138613861384e-06,
+      "loss": 0.0026,
+      "step": 692
+    },
+    {
+      "epoch": 3.4012269938650306,
+      "grad_norm": 0.02160765416920185,
+      "learning_rate": 2.2316831683168315e-06,
+      "loss": 0.0014,
+      "step": 693
+    },
+    {
+      "epoch": 3.4061349693251532,
+      "grad_norm": 0.012106803245842457,
+      "learning_rate": 2.2247524752475246e-06,
+      "loss": 0.0008,
+      "step": 694
+    },
+    {
+      "epoch": 3.411042944785276,
+      "grad_norm": 0.027164770290255547,
+      "learning_rate": 2.2178217821782176e-06,
+      "loss": 0.001,
+      "step": 695
+    },
+    {
+      "epoch": 3.4159509202453986,
+      "grad_norm": 0.03465467691421509,
+      "learning_rate": 2.2108910891089107e-06,
+      "loss": 0.0008,
+      "step": 696
+    },
+    {
+      "epoch": 3.4208588957055213,
+      "grad_norm": 0.01086588017642498,
+      "learning_rate": 2.2039603960396038e-06,
+      "loss": 0.0016,
+      "step": 697
+    },
+    {
+      "epoch": 3.425766871165644,
+      "grad_norm": 0.2833847105503082,
+      "learning_rate": 2.197029702970297e-06,
+      "loss": 0.0092,
+      "step": 698
+    },
+    {
+      "epoch": 3.430674846625767,
+      "grad_norm": 0.01620599813759327,
+      "learning_rate": 2.19009900990099e-06,
+      "loss": 0.0005,
+      "step": 699
+    },
+    {
+      "epoch": 3.4355828220858897,
+      "grad_norm": 0.0964425802230835,
+      "learning_rate": 2.183168316831683e-06,
+      "loss": 0.0059,
+      "step": 700
+    },
+    {
+      "epoch": 3.4404907975460124,
+      "grad_norm": 0.051153287291526794,
+      "learning_rate": 2.176237623762376e-06,
+      "loss": 0.003,
+      "step": 701
+    },
+    {
+      "epoch": 3.445398773006135,
+      "grad_norm": 0.12504975497722626,
+      "learning_rate": 2.169306930693069e-06,
+      "loss": 0.0023,
+      "step": 702
+    },
+    {
+      "epoch": 3.4503067484662577,
+      "grad_norm": 0.18094071745872498,
+      "learning_rate": 2.162376237623762e-06,
+      "loss": 0.0175,
+      "step": 703
+    },
+    {
+      "epoch": 3.4552147239263804,
+      "grad_norm": 0.011514030396938324,
+      "learning_rate": 2.1554455445544553e-06,
+      "loss": 0.001,
+      "step": 704
+    },
+    {
+      "epoch": 3.460122699386503,
+      "grad_norm": 0.015152939595282078,
+      "learning_rate": 2.1485148514851483e-06,
+      "loss": 0.0011,
+      "step": 705
+    },
+    {
+      "epoch": 3.4650306748466257,
+      "grad_norm": 0.05039620399475098,
+      "learning_rate": 2.1415841584158414e-06,
+      "loss": 0.0049,
+      "step": 706
+    },
+    {
+      "epoch": 3.4699386503067484,
+      "grad_norm": 0.044066257774829865,
+      "learning_rate": 2.1346534653465345e-06,
+      "loss": 0.0011,
+      "step": 707
+    },
+    {
+      "epoch": 3.474846625766871,
+      "grad_norm": 0.06301417946815491,
+      "learning_rate": 2.1277227722772275e-06,
+      "loss": 0.002,
+      "step": 708
+    },
+    {
+      "epoch": 3.4797546012269938,
+      "grad_norm": 0.05275435373187065,
+      "learning_rate": 2.1207920792079206e-06,
+      "loss": 0.0026,
+      "step": 709
+    },
+    {
+      "epoch": 3.4846625766871164,
+      "grad_norm": 0.05170956999063492,
+      "learning_rate": 2.1138613861386137e-06,
+      "loss": 0.0033,
+      "step": 710
+    },
+    {
+      "epoch": 3.489570552147239,
+      "grad_norm": 0.12438485026359558,
+      "learning_rate": 2.1069306930693067e-06,
+      "loss": 0.0028,
+      "step": 711
+    },
+    {
+      "epoch": 3.4944785276073618,
+      "grad_norm": 0.07120586186647415,
+      "learning_rate": 2.1e-06,
+      "loss": 0.0021,
+      "step": 712
+    },
+    {
+      "epoch": 3.499386503067485,
+      "grad_norm": 0.027411244809627533,
+      "learning_rate": 2.093069306930693e-06,
+      "loss": 0.0011,
+      "step": 713
+    },
+    {
+      "epoch": 3.5042944785276076,
+      "grad_norm": 0.005563246086239815,
+      "learning_rate": 2.086138613861386e-06,
+      "loss": 0.0007,
+      "step": 714
+    },
+    {
+      "epoch": 3.5092024539877302,
+      "grad_norm": 0.04439758136868477,
+      "learning_rate": 2.079207920792079e-06,
+      "loss": 0.0017,
+      "step": 715
+    },
+    {
+      "epoch": 3.514110429447853,
+      "grad_norm": 0.04545675963163376,
+      "learning_rate": 2.072277227722772e-06,
+      "loss": 0.0045,
+      "step": 716
+    },
+    {
+      "epoch": 3.5190184049079756,
+      "grad_norm": 0.019012991338968277,
+      "learning_rate": 2.065346534653465e-06,
+      "loss": 0.001,
+      "step": 717
+    },
+    {
+      "epoch": 3.5239263803680982,
+      "grad_norm": 0.5096023678779602,
+      "learning_rate": 2.058415841584158e-06,
+      "loss": 0.0205,
+      "step": 718
+    },
+    {
+      "epoch": 3.528834355828221,
+      "grad_norm": 0.03077244944870472,
+      "learning_rate": 2.0514851485148513e-06,
+      "loss": 0.0018,
+      "step": 719
+    },
+    {
+      "epoch": 3.5337423312883436,
+      "grad_norm": 0.052647169679403305,
+      "learning_rate": 2.0445544554455443e-06,
+      "loss": 0.0021,
+      "step": 720
+    },
+    {
+      "epoch": 3.5386503067484663,
+      "grad_norm": 0.014248156920075417,
+      "learning_rate": 2.0376237623762374e-06,
+      "loss": 0.0007,
+      "step": 721
+    },
+    {
+      "epoch": 3.543558282208589,
+      "grad_norm": 0.11832743138074875,
+      "learning_rate": 2.0306930693069305e-06,
+      "loss": 0.0088,
+      "step": 722
+    },
+    {
+      "epoch": 3.5484662576687116,
+      "grad_norm": 0.025962911546230316,
+      "learning_rate": 2.0237623762376236e-06,
+      "loss": 0.0009,
+      "step": 723
+    },
+    {
+      "epoch": 3.5533742331288343,
+      "grad_norm": 0.06493301689624786,
+      "learning_rate": 2.0168316831683166e-06,
+      "loss": 0.0024,
+      "step": 724
+    },
+    {
+      "epoch": 3.558282208588957,
+      "grad_norm": 0.023671971634030342,
+      "learning_rate": 2.0099009900990097e-06,
+      "loss": 0.0019,
+      "step": 725
+    },
+    {
+      "epoch": 3.5631901840490796,
+      "grad_norm": 0.02273421734571457,
+      "learning_rate": 2.0029702970297028e-06,
+      "loss": 0.0008,
+      "step": 726
+    },
+    {
+      "epoch": 3.5680981595092023,
+      "grad_norm": 0.18828389048576355,
+      "learning_rate": 1.996039603960396e-06,
+      "loss": 0.0182,
+      "step": 727
+    },
+    {
+      "epoch": 3.573006134969325,
+      "grad_norm": 0.01973796635866165,
+      "learning_rate": 1.989108910891089e-06,
+      "loss": 0.0023,
+      "step": 728
+    },
+    {
+      "epoch": 3.5779141104294476,
+      "grad_norm": 0.03295096009969711,
+      "learning_rate": 1.982178217821782e-06,
+      "loss": 0.0017,
+      "step": 729
+    },
+    {
+      "epoch": 3.5828220858895703,
+      "grad_norm": 0.012010748498141766,
+      "learning_rate": 1.975247524752475e-06,
+      "loss": 0.0009,
+      "step": 730
+    },
+    {
+      "epoch": 3.5877300613496934,
+      "grad_norm": 0.008494194597005844,
+      "learning_rate": 1.9683168316831685e-06,
+      "loss": 0.0009,
+      "step": 731
+    },
+    {
+      "epoch": 3.592638036809816,
+      "grad_norm": 0.016253001987934113,
+      "learning_rate": 1.9613861386138616e-06,
+      "loss": 0.0015,
+      "step": 732
+    },
+    {
+      "epoch": 3.5975460122699388,
+      "grad_norm": 0.007456593681126833,
+      "learning_rate": 1.9544554455445547e-06,
+      "loss": 0.0007,
+      "step": 733
+    },
+    {
+      "epoch": 3.6024539877300614,
+      "grad_norm": 0.00861444789916277,
+      "learning_rate": 1.9475247524752477e-06,
+      "loss": 0.0007,
+      "step": 734
+    },
+    {
+      "epoch": 3.607361963190184,
+      "grad_norm": 0.0065794652327895164,
+      "learning_rate": 1.940594059405941e-06,
+      "loss": 0.001,
+      "step": 735
+    },
+    {
+      "epoch": 3.6122699386503068,
+      "grad_norm": 0.015389169566333294,
+      "learning_rate": 1.933663366336634e-06,
+      "loss": 0.0015,
+      "step": 736
+    },
+    {
+      "epoch": 3.6171779141104294,
+      "grad_norm": 0.025337016209959984,
+      "learning_rate": 1.926732673267327e-06,
+      "loss": 0.0018,
+      "step": 737
+    },
+    {
+      "epoch": 3.622085889570552,
+      "grad_norm": 0.00653579318895936,
+      "learning_rate": 1.91980198019802e-06,
+      "loss": 0.0008,
+      "step": 738
+    },
+    {
+      "epoch": 3.626993865030675,
+      "grad_norm": 0.10997878760099411,
+      "learning_rate": 1.912871287128713e-06,
+      "loss": 0.0068,
+      "step": 739
+    },
+    {
+      "epoch": 3.6319018404907975,
+      "grad_norm": 0.05580228194594383,
+      "learning_rate": 1.9059405940594061e-06,
+      "loss": 0.0056,
+      "step": 740
+    },
+    {
+      "epoch": 3.63680981595092,
+      "grad_norm": 0.06799723207950592,
+      "learning_rate": 1.8990099009900992e-06,
+      "loss": 0.0019,
+      "step": 741
+    },
+    {
+      "epoch": 3.641717791411043,
+      "grad_norm": 0.20822834968566895,
+      "learning_rate": 1.8920792079207923e-06,
+      "loss": 0.0039,
+      "step": 742
+    },
+    {
+      "epoch": 3.646625766871166,
+      "grad_norm": 0.03920517861843109,
+      "learning_rate": 1.8851485148514851e-06,
+      "loss": 0.0017,
+      "step": 743
+    },
+    {
+      "epoch": 3.6515337423312886,
+      "grad_norm": 0.06821847707033157,
+      "learning_rate": 1.8782178217821782e-06,
+      "loss": 0.0027,
+      "step": 744
+    },
+    {
+      "epoch": 3.6564417177914113,
+      "grad_norm": 0.09687570482492447,
+      "learning_rate": 1.8712871287128713e-06,
+      "loss": 0.002,
+      "step": 745
+    },
+    {
+      "epoch": 3.661349693251534,
+      "grad_norm": 0.05403744429349899,
+      "learning_rate": 1.8643564356435643e-06,
+      "loss": 0.0019,
+      "step": 746
+    },
+    {
+      "epoch": 3.6662576687116566,
+      "grad_norm": 0.019597845152020454,
+      "learning_rate": 1.8574257425742574e-06,
+      "loss": 0.0009,
+      "step": 747
+    },
+    {
+      "epoch": 3.6711656441717793,
+      "grad_norm": 0.04923088103532791,
+      "learning_rate": 1.8504950495049505e-06,
+      "loss": 0.0013,
+      "step": 748
+    },
+    {
+      "epoch": 3.676073619631902,
+      "grad_norm": 0.0967707633972168,
+      "learning_rate": 1.8435643564356435e-06,
+      "loss": 0.0039,
+      "step": 749
+    },
+    {
+      "epoch": 3.6809815950920246,
+      "grad_norm": 0.037127815186977386,
+      "learning_rate": 1.8366336633663366e-06,
+      "loss": 0.0012,
+      "step": 750
+    },
+    {
+      "epoch": 3.6858895705521473,
+      "grad_norm": 0.013236461207270622,
+      "learning_rate": 1.8297029702970297e-06,
+      "loss": 0.0045,
+      "step": 751
+    },
+    {
+      "epoch": 3.69079754601227,
+      "grad_norm": 0.031925372779369354,
+      "learning_rate": 1.8227722772277228e-06,
+      "loss": 0.002,
+      "step": 752
+    },
+    {
+      "epoch": 3.6957055214723926,
+      "grad_norm": 0.023648735135793686,
+      "learning_rate": 1.8158415841584158e-06,
+      "loss": 0.0012,
+      "step": 753
+    },
+    {
+      "epoch": 3.7006134969325153,
+      "grad_norm": 0.01484636776149273,
+      "learning_rate": 1.8089108910891089e-06,
+      "loss": 0.0011,
+      "step": 754
+    },
+    {
+      "epoch": 3.705521472392638,
+      "grad_norm": 0.02288316749036312,
+      "learning_rate": 1.801980198019802e-06,
+      "loss": 0.0018,
+      "step": 755
+    },
+    {
+      "epoch": 3.7104294478527606,
+      "grad_norm": 0.005614751018583775,
+      "learning_rate": 1.795049504950495e-06,
+      "loss": 0.001,
+      "step": 756
+    },
+    {
+      "epoch": 3.7153374233128833,
+      "grad_norm": 0.03587134927511215,
+      "learning_rate": 1.788118811881188e-06,
+      "loss": 0.0013,
+      "step": 757
+    },
+    {
+      "epoch": 3.720245398773006,
+      "grad_norm": 0.048482466489076614,
+      "learning_rate": 1.7811881188118812e-06,
+      "loss": 0.0014,
+      "step": 758
+    },
+    {
+      "epoch": 3.7251533742331286,
+      "grad_norm": 0.06541978567838669,
+      "learning_rate": 1.7742574257425742e-06,
+      "loss": 0.0023,
+      "step": 759
+    },
+    {
+      "epoch": 3.7300613496932513,
+      "grad_norm": 0.040501050651073456,
+      "learning_rate": 1.7673267326732673e-06,
+      "loss": 0.0014,
+      "step": 760
+    },
+    {
+      "epoch": 3.734969325153374,
+      "grad_norm": 0.006551014259457588,
+      "learning_rate": 1.7603960396039604e-06,
+      "loss": 0.0009,
+      "step": 761
+    },
+    {
+      "epoch": 3.7398773006134967,
+      "grad_norm": 0.11849401146173477,
+      "learning_rate": 1.7534653465346534e-06,
+      "loss": 0.0099,
+      "step": 762
+    },
+    {
+      "epoch": 3.7447852760736198,
+      "grad_norm": 0.004786093719303608,
+      "learning_rate": 1.7465346534653465e-06,
+      "loss": 0.0008,
+      "step": 763
+    },
+    {
+      "epoch": 3.7496932515337424,
+      "grad_norm": 0.02577151544392109,
+      "learning_rate": 1.7396039603960396e-06,
+      "loss": 0.0016,
+      "step": 764
+    },
+    {
+      "epoch": 3.754601226993865,
+      "grad_norm": 0.014097603037953377,
+      "learning_rate": 1.7326732673267326e-06,
+      "loss": 0.0008,
+      "step": 765
+    },
+    {
+      "epoch": 3.759509202453988,
+      "grad_norm": 0.05258313938975334,
+      "learning_rate": 1.7257425742574257e-06,
+      "loss": 0.0009,
+      "step": 766
+    },
+    {
+      "epoch": 3.7644171779141105,
+      "grad_norm": 0.09022804349660873,
+      "learning_rate": 1.7188118811881188e-06,
+      "loss": 0.005,
+      "step": 767
+    },
+    {
+      "epoch": 3.769325153374233,
+      "grad_norm": 0.008886247873306274,
+      "learning_rate": 1.7118811881188119e-06,
+      "loss": 0.0008,
+      "step": 768
+    },
+    {
+      "epoch": 3.774233128834356,
+      "grad_norm": 0.036997053772211075,
+      "learning_rate": 1.704950495049505e-06,
+      "loss": 0.0011,
+      "step": 769
+    },
+    {
+      "epoch": 3.7791411042944785,
+      "grad_norm": 0.05569405481219292,
+      "learning_rate": 1.698019801980198e-06,
+      "loss": 0.0018,
+      "step": 770
+    },
+    {
+      "epoch": 3.784049079754601,
+      "grad_norm": 0.0031505110673606396,
+      "learning_rate": 1.691089108910891e-06,
+      "loss": 0.0004,
+      "step": 771
+    },
+    {
+      "epoch": 3.788957055214724,
+      "grad_norm": 0.014605509117245674,
+      "learning_rate": 1.6841584158415841e-06,
+      "loss": 0.0011,
+      "step": 772
+    },
+    {
+      "epoch": 3.7938650306748465,
+      "grad_norm": 0.09325973689556122,
+      "learning_rate": 1.6772277227722772e-06,
+      "loss": 0.0052,
+      "step": 773
+    },
+    {
+      "epoch": 3.7987730061349696,
+      "grad_norm": 0.059272442013025284,
+      "learning_rate": 1.6702970297029703e-06,
+      "loss": 0.0022,
+      "step": 774
+    },
+    {
+      "epoch": 3.8036809815950923,
+      "grad_norm": 0.01452575996518135,
+      "learning_rate": 1.6633663366336633e-06,
+      "loss": 0.0021,
+      "step": 775
+    },
+    {
+      "epoch": 3.808588957055215,
+      "grad_norm": 0.17578046023845673,
+      "learning_rate": 1.6564356435643564e-06,
+      "loss": 0.0093,
+      "step": 776
+    },
+    {
+      "epoch": 3.8134969325153376,
+      "grad_norm": 0.007930277846753597,
+      "learning_rate": 1.6495049504950495e-06,
+      "loss": 0.0006,
+      "step": 777
+    },
+    {
+      "epoch": 3.8184049079754603,
+      "grad_norm": 0.07230112701654434,
+      "learning_rate": 1.6425742574257425e-06,
+      "loss": 0.0025,
+      "step": 778
+    },
+    {
+      "epoch": 3.823312883435583,
+      "grad_norm": 0.03507319092750549,
+      "learning_rate": 1.6356435643564358e-06,
+      "loss": 0.0017,
+      "step": 779
+    },
+    {
+      "epoch": 3.8282208588957056,
+      "grad_norm": 0.06336654722690582,
+      "learning_rate": 1.6287128712871287e-06,
+      "loss": 0.0025,
+      "step": 780
+    },
+    {
+      "epoch": 3.8331288343558283,
+      "grad_norm": 0.14077608287334442,
+      "learning_rate": 1.6217821782178217e-06,
+      "loss": 0.0086,
+      "step": 781
+    },
+    {
+      "epoch": 3.838036809815951,
+      "grad_norm": 0.015772581100463867,
+      "learning_rate": 1.6148514851485148e-06,
+      "loss": 0.001,
+      "step": 782
+    },
+    {
+      "epoch": 3.8429447852760736,
+      "grad_norm": 0.01927962154150009,
+      "learning_rate": 1.6079207920792079e-06,
+      "loss": 0.0015,
+      "step": 783
+    },
+    {
+      "epoch": 3.8478527607361963,
+      "grad_norm": 0.011015449650585651,
+      "learning_rate": 1.600990099009901e-06,
+      "loss": 0.0008,
+      "step": 784
+    },
+    {
+      "epoch": 3.852760736196319,
+      "grad_norm": 0.40098482370376587,
+      "learning_rate": 1.594059405940594e-06,
+      "loss": 0.0314,
+      "step": 785
+    },
+    {
+      "epoch": 3.8576687116564417,
+      "grad_norm": 0.02672453783452511,
+      "learning_rate": 1.587128712871287e-06,
+      "loss": 0.0025,
+      "step": 786
+    },
+    {
+      "epoch": 3.8625766871165643,
+      "grad_norm": 0.022412395104765892,
+      "learning_rate": 1.5801980198019802e-06,
+      "loss": 0.0013,
+      "step": 787
+    },
+    {
+      "epoch": 3.867484662576687,
+      "grad_norm": 0.023978037759661674,
+      "learning_rate": 1.5732673267326732e-06,
+      "loss": 0.0017,
+      "step": 788
+    },
+    {
+      "epoch": 3.8723926380368097,
+      "grad_norm": 0.017764659598469734,
+      "learning_rate": 1.5663366336633663e-06,
+      "loss": 0.002,
+      "step": 789
+    },
+    {
+      "epoch": 3.8773006134969323,
+      "grad_norm": 0.012586713768541813,
+      "learning_rate": 1.5594059405940594e-06,
+      "loss": 0.0008,
+      "step": 790
+    },
+    {
+      "epoch": 3.882208588957055,
+      "grad_norm": 0.056462038308382034,
+      "learning_rate": 1.5524752475247524e-06,
+      "loss": 0.0036,
+      "step": 791
+    },
+    {
+      "epoch": 3.8871165644171777,
+      "grad_norm": 0.05329478159546852,
+      "learning_rate": 1.5455445544554455e-06,
+      "loss": 0.0041,
+      "step": 792
+    },
+    {
+      "epoch": 3.8920245398773003,
+      "grad_norm": 0.0013215028448030353,
+      "learning_rate": 1.5386138613861386e-06,
+      "loss": 0.0004,
+      "step": 793
+    },
+    {
+      "epoch": 3.8969325153374235,
+      "grad_norm": 0.05318621173501015,
+      "learning_rate": 1.5316831683168316e-06,
+      "loss": 0.0011,
+      "step": 794
+    },
+    {
+      "epoch": 3.901840490797546,
+      "grad_norm": 0.3169184625148773,
+      "learning_rate": 1.5247524752475247e-06,
+      "loss": 0.0221,
+      "step": 795
+    },
+    {
+      "epoch": 3.906748466257669,
+      "grad_norm": 0.04726627469062805,
+      "learning_rate": 1.5178217821782178e-06,
+      "loss": 0.0015,
+      "step": 796
+    },
+    {
+      "epoch": 3.9116564417177915,
+      "grad_norm": 0.13995185494422913,
+      "learning_rate": 1.5108910891089108e-06,
+      "loss": 0.0092,
+      "step": 797
+    },
+    {
+      "epoch": 3.916564417177914,
+      "grad_norm": 0.01544391643255949,
+      "learning_rate": 1.503960396039604e-06,
+      "loss": 0.0012,
+      "step": 798
+    },
+    {
+      "epoch": 3.921472392638037,
+      "grad_norm": 0.1588226556777954,
+      "learning_rate": 1.497029702970297e-06,
+      "loss": 0.0087,
+      "step": 799
+    },
+    {
+      "epoch": 3.9263803680981595,
+      "grad_norm": 0.011546803638339043,
+      "learning_rate": 1.49009900990099e-06,
+      "loss": 0.0015,
+      "step": 800
+    },
+    {
+      "epoch": 3.931288343558282,
+      "grad_norm": 0.04798766225576401,
+      "learning_rate": 1.4831683168316831e-06,
+      "loss": 0.0013,
+      "step": 801
+    },
+    {
+      "epoch": 3.936196319018405,
+      "grad_norm": 0.01064328383654356,
+      "learning_rate": 1.4762376237623762e-06,
+      "loss": 0.0011,
+      "step": 802
+    },
+    {
+      "epoch": 3.9411042944785275,
+      "grad_norm": 0.1379479169845581,
+      "learning_rate": 1.4693069306930693e-06,
+      "loss": 0.0142,
+      "step": 803
+    },
+    {
+      "epoch": 3.94601226993865,
+      "grad_norm": 0.054966770112514496,
+      "learning_rate": 1.4623762376237623e-06,
+      "loss": 0.0014,
+      "step": 804
+    },
+    {
+      "epoch": 3.950920245398773,
+      "grad_norm": 0.035458799451589584,
+      "learning_rate": 1.4554455445544554e-06,
+      "loss": 0.002,
+      "step": 805
+    },
+    {
+      "epoch": 3.955828220858896,
+      "grad_norm": 0.011258352547883987,
+      "learning_rate": 1.4485148514851485e-06,
+      "loss": 0.0005,
+      "step": 806
+    },
+    {
+      "epoch": 3.9607361963190186,
+      "grad_norm": 0.022768640890717506,
+      "learning_rate": 1.4415841584158415e-06,
+      "loss": 0.0009,
+      "step": 807
+    },
+    {
+      "epoch": 3.9656441717791413,
+      "grad_norm": 0.0772656723856926,
+      "learning_rate": 1.4346534653465346e-06,
+      "loss": 0.0014,
+      "step": 808
+    },
+    {
+      "epoch": 3.970552147239264,
+      "grad_norm": 0.06587695330381393,
+      "learning_rate": 1.4277227722772277e-06,
+      "loss": 0.0034,
+      "step": 809
+    },
+    {
+      "epoch": 3.9754601226993866,
+      "grad_norm": 0.01118537038564682,
+      "learning_rate": 1.420792079207921e-06,
+      "loss": 0.0009,
+      "step": 810
+    },
+    {
+      "epoch": 3.9803680981595093,
+      "grad_norm": 0.06560896337032318,
+      "learning_rate": 1.413861386138614e-06,
+      "loss": 0.0019,
+      "step": 811
+    },
+    {
+      "epoch": 3.985276073619632,
+      "grad_norm": 0.014048455283045769,
+      "learning_rate": 1.406930693069307e-06,
+      "loss": 0.0018,
+      "step": 812
+    },
+    {
+      "epoch": 3.9901840490797547,
+      "grad_norm": 0.01656423695385456,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.001,
+      "step": 813
+    },
+    {
+      "epoch": 3.9950920245398773,
+      "grad_norm": 0.0036234341096132994,
+      "learning_rate": 1.3930693069306932e-06,
+      "loss": 0.0007,
+      "step": 814
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.012785837985575199,
+      "learning_rate": 1.3861386138613863e-06,
+      "loss": 0.0009,
+      "step": 815
+    },
+    {
+      "epoch": 4.004907975460123,
+      "grad_norm": 0.025322729721665382,
+      "learning_rate": 1.3792079207920791e-06,
+      "loss": 0.002,
+      "step": 816
+    },
+    {
+      "epoch": 4.009815950920245,
+      "grad_norm": 0.2052641361951828,
+      "learning_rate": 1.3722772277227722e-06,
+      "loss": 0.0056,
+      "step": 817
+    },
+    {
+      "epoch": 4.014723926380368,
+      "grad_norm": 0.057693980634212494,
+      "learning_rate": 1.3653465346534653e-06,
+      "loss": 0.0021,
+      "step": 818
+    },
+    {
+      "epoch": 4.019631901840491,
+      "grad_norm": 0.009920844808220863,
+      "learning_rate": 1.3584158415841583e-06,
+      "loss": 0.0009,
+      "step": 819
+    },
+    {
+      "epoch": 4.024539877300613,
+      "grad_norm": 0.18843849003314972,
+      "learning_rate": 1.3514851485148514e-06,
+      "loss": 0.0148,
+      "step": 820
+    },
+    {
+      "epoch": 4.029447852760736,
+      "grad_norm": 0.003511168295517564,
+      "learning_rate": 1.3445544554455445e-06,
+      "loss": 0.0003,
+      "step": 821
+    },
+    {
+      "epoch": 4.034355828220859,
+      "grad_norm": 0.02023676596581936,
+      "learning_rate": 1.3376237623762376e-06,
+      "loss": 0.001,
+      "step": 822
+    },
+    {
+      "epoch": 4.039263803680981,
+      "grad_norm": 0.010772217065095901,
+      "learning_rate": 1.3306930693069306e-06,
+      "loss": 0.0013,
+      "step": 823
+    },
+    {
+      "epoch": 4.044171779141104,
+      "grad_norm": 0.023414717987179756,
+      "learning_rate": 1.3237623762376237e-06,
+      "loss": 0.0023,
+      "step": 824
+    },
+    {
+      "epoch": 4.049079754601227,
+      "grad_norm": 0.019114743918180466,
+      "learning_rate": 1.3168316831683168e-06,
+      "loss": 0.0007,
+      "step": 825
+    },
+    {
+      "epoch": 4.053987730061349,
+      "grad_norm": 0.012856281362473965,
+      "learning_rate": 1.3099009900990098e-06,
+      "loss": 0.0015,
+      "step": 826
+    },
+    {
+      "epoch": 4.058895705521472,
+      "grad_norm": 0.00855772290378809,
+      "learning_rate": 1.302970297029703e-06,
+      "loss": 0.0015,
+      "step": 827
+    },
+    {
+      "epoch": 4.063803680981595,
+      "grad_norm": 0.021148694679141045,
+      "learning_rate": 1.296039603960396e-06,
+      "loss": 0.0012,
+      "step": 828
+    },
+    {
+      "epoch": 4.068711656441717,
+      "grad_norm": 0.07430653274059296,
+      "learning_rate": 1.289108910891089e-06,
+      "loss": 0.0049,
+      "step": 829
+    },
+    {
+      "epoch": 4.07361963190184,
+      "grad_norm": 0.10033933073282242,
+      "learning_rate": 1.282178217821782e-06,
+      "loss": 0.0026,
+      "step": 830
+    },
+    {
+      "epoch": 4.078527607361964,
+      "grad_norm": 0.006094958167523146,
+      "learning_rate": 1.2752475247524752e-06,
+      "loss": 0.0006,
+      "step": 831
+    },
+    {
+      "epoch": 4.083435582822086,
+      "grad_norm": 0.06336677074432373,
+      "learning_rate": 1.2683168316831682e-06,
+      "loss": 0.0038,
+      "step": 832
+    },
+    {
+      "epoch": 4.088343558282209,
+      "grad_norm": 0.005283738486468792,
+      "learning_rate": 1.2613861386138613e-06,
+      "loss": 0.0015,
+      "step": 833
+    },
+    {
+      "epoch": 4.093251533742332,
+      "grad_norm": 0.006675936747342348,
+      "learning_rate": 1.2544554455445544e-06,
+      "loss": 0.0006,
+      "step": 834
+    },
+    {
+      "epoch": 4.098159509202454,
+      "grad_norm": 0.008660698309540749,
+      "learning_rate": 1.2475247524752474e-06,
+      "loss": 0.0018,
+      "step": 835
+    },
+    {
+      "epoch": 4.103067484662577,
+      "grad_norm": 0.02305518463253975,
+      "learning_rate": 1.2405940594059405e-06,
+      "loss": 0.0007,
+      "step": 836
+    },
+    {
+      "epoch": 4.1079754601227,
+      "grad_norm": 0.024816259741783142,
+      "learning_rate": 1.2336633663366336e-06,
+      "loss": 0.0007,
+      "step": 837
+    },
+    {
+      "epoch": 4.112883435582822,
+      "grad_norm": 0.026276560500264168,
+      "learning_rate": 1.2267326732673267e-06,
+      "loss": 0.0016,
+      "step": 838
+    },
+    {
+      "epoch": 4.117791411042945,
+      "grad_norm": 0.029642153531312943,
+      "learning_rate": 1.2198019801980197e-06,
+      "loss": 0.0038,
+      "step": 839
+    },
+    {
+      "epoch": 4.122699386503068,
+      "grad_norm": 0.05285963416099548,
+      "learning_rate": 1.2128712871287128e-06,
+      "loss": 0.0019,
+      "step": 840
+    },
+    {
+      "epoch": 4.12760736196319,
+      "grad_norm": 0.014620939269661903,
+      "learning_rate": 1.2059405940594059e-06,
+      "loss": 0.0012,
+      "step": 841
+    },
+    {
+      "epoch": 4.132515337423313,
+      "grad_norm": 0.012532511726021767,
+      "learning_rate": 1.1990099009900991e-06,
+      "loss": 0.0007,
+      "step": 842
+    },
+    {
+      "epoch": 4.137423312883436,
+      "grad_norm": 0.04022945091128349,
+      "learning_rate": 1.1920792079207922e-06,
+      "loss": 0.0032,
+      "step": 843
+    },
+    {
+      "epoch": 4.142331288343558,
+      "grad_norm": 0.021105729043483734,
+      "learning_rate": 1.1851485148514853e-06,
+      "loss": 0.0015,
+      "step": 844
+    },
+    {
+      "epoch": 4.147239263803681,
+      "grad_norm": 0.07277761399745941,
+      "learning_rate": 1.1782178217821783e-06,
+      "loss": 0.0038,
+      "step": 845
+    },
+    {
+      "epoch": 4.152147239263804,
+      "grad_norm": 0.015429302118718624,
+      "learning_rate": 1.1712871287128714e-06,
+      "loss": 0.0014,
+      "step": 846
+    },
+    {
+      "epoch": 4.157055214723926,
+      "grad_norm": 0.02602989971637726,
+      "learning_rate": 1.1643564356435645e-06,
+      "loss": 0.0013,
+      "step": 847
+    },
+    {
+      "epoch": 4.161963190184049,
+      "grad_norm": 0.018687183037400246,
+      "learning_rate": 1.1574257425742575e-06,
+      "loss": 0.0016,
+      "step": 848
+    },
+    {
+      "epoch": 4.166871165644172,
+      "grad_norm": 0.019744986668229103,
+      "learning_rate": 1.1504950495049506e-06,
+      "loss": 0.0013,
+      "step": 849
+    },
+    {
+      "epoch": 4.171779141104294,
+      "grad_norm": 0.029573217034339905,
+      "learning_rate": 1.1435643564356437e-06,
+      "loss": 0.0023,
+      "step": 850
+    },
+    {
+      "epoch": 4.176687116564417,
+      "grad_norm": 0.020479142665863037,
+      "learning_rate": 1.1366336633663368e-06,
+      "loss": 0.0011,
+      "step": 851
+    },
+    {
+      "epoch": 4.18159509202454,
+      "grad_norm": 0.11432457715272903,
+      "learning_rate": 1.1297029702970298e-06,
+      "loss": 0.0041,
+      "step": 852
+    },
+    {
+      "epoch": 4.186503067484662,
+      "grad_norm": 0.012511249631643295,
+      "learning_rate": 1.1227722772277227e-06,
+      "loss": 0.0006,
+      "step": 853
+    },
+    {
+      "epoch": 4.191411042944785,
+      "grad_norm": 0.01595146209001541,
+      "learning_rate": 1.1158415841584157e-06,
+      "loss": 0.0009,
+      "step": 854
+    },
+    {
+      "epoch": 4.196319018404908,
+      "grad_norm": 0.007092094514518976,
+      "learning_rate": 1.1089108910891088e-06,
+      "loss": 0.0007,
+      "step": 855
+    },
+    {
+      "epoch": 4.20122699386503,
+      "grad_norm": 0.03979247063398361,
+      "learning_rate": 1.1019801980198019e-06,
+      "loss": 0.0018,
+      "step": 856
+    },
+    {
+      "epoch": 4.206134969325153,
+      "grad_norm": 0.008829467929899693,
+      "learning_rate": 1.095049504950495e-06,
+      "loss": 0.0007,
+      "step": 857
+    },
+    {
+      "epoch": 4.211042944785276,
+      "grad_norm": 0.09763351082801819,
+      "learning_rate": 1.088118811881188e-06,
+      "loss": 0.0048,
+      "step": 858
+    },
+    {
+      "epoch": 4.215950920245398,
+      "grad_norm": 0.08337781578302383,
+      "learning_rate": 1.081188118811881e-06,
+      "loss": 0.0014,
+      "step": 859
+    },
+    {
+      "epoch": 4.220858895705521,
+      "grad_norm": 0.029353009536862373,
+      "learning_rate": 1.0742574257425742e-06,
+      "loss": 0.001,
+      "step": 860
+    },
+    {
+      "epoch": 4.225766871165644,
+      "grad_norm": 0.121429443359375,
+      "learning_rate": 1.0673267326732672e-06,
+      "loss": 0.0148,
+      "step": 861
+    },
+    {
+      "epoch": 4.230674846625767,
+      "grad_norm": 0.01580023020505905,
+      "learning_rate": 1.0603960396039603e-06,
+      "loss": 0.001,
+      "step": 862
+    },
+    {
+      "epoch": 4.23558282208589,
+      "grad_norm": 0.013746123760938644,
+      "learning_rate": 1.0534653465346534e-06,
+      "loss": 0.0012,
+      "step": 863
+    },
+    {
+      "epoch": 4.240490797546013,
+      "grad_norm": 0.011870300397276878,
+      "learning_rate": 1.0465346534653464e-06,
+      "loss": 0.0007,
+      "step": 864
+    },
+    {
+      "epoch": 4.245398773006135,
+      "grad_norm": 0.05088931694626808,
+      "learning_rate": 1.0396039603960395e-06,
+      "loss": 0.0017,
+      "step": 865
+    },
+    {
+      "epoch": 4.250306748466258,
+      "grad_norm": 0.38851794600486755,
+      "learning_rate": 1.0326732673267326e-06,
+      "loss": 0.014,
+      "step": 866
+    },
+    {
+      "epoch": 4.255214723926381,
+      "grad_norm": 0.01347925141453743,
+      "learning_rate": 1.0257425742574256e-06,
+      "loss": 0.0012,
+      "step": 867
+    },
+    {
+      "epoch": 4.260122699386503,
+      "grad_norm": 0.010402753949165344,
+      "learning_rate": 1.0188118811881187e-06,
+      "loss": 0.0007,
+      "step": 868
+    },
+    {
+      "epoch": 4.265030674846626,
+      "grad_norm": 0.03338263928890228,
+      "learning_rate": 1.0118811881188118e-06,
+      "loss": 0.0017,
+      "step": 869
+    },
+    {
+      "epoch": 4.269938650306749,
+      "grad_norm": 0.007806051056832075,
+      "learning_rate": 1.0049504950495048e-06,
+      "loss": 0.0006,
+      "step": 870
+    },
+    {
+      "epoch": 4.274846625766871,
+      "grad_norm": 0.0339755155146122,
+      "learning_rate": 9.98019801980198e-07,
+      "loss": 0.0027,
+      "step": 871
+    },
+    {
+      "epoch": 4.279754601226994,
+      "grad_norm": 0.02654801867902279,
+      "learning_rate": 9.91089108910891e-07,
+      "loss": 0.0012,
+      "step": 872
+    },
+    {
+      "epoch": 4.284662576687117,
+      "grad_norm": 0.018412116914987564,
+      "learning_rate": 9.841584158415843e-07,
+      "loss": 0.0022,
+      "step": 873
+    },
+    {
+      "epoch": 4.289570552147239,
+      "grad_norm": 0.0640820562839508,
+      "learning_rate": 9.772277227722773e-07,
+      "loss": 0.0021,
+      "step": 874
+    },
+    {
+      "epoch": 4.294478527607362,
+      "grad_norm": 0.03333529084920883,
+      "learning_rate": 9.702970297029704e-07,
+      "loss": 0.0023,
+      "step": 875
+    },
+    {
+      "epoch": 4.299386503067485,
+      "grad_norm": 0.022033028304576874,
+      "learning_rate": 9.633663366336635e-07,
+      "loss": 0.0007,
+      "step": 876
+    },
+    {
+      "epoch": 4.304294478527607,
+      "grad_norm": 0.010194915346801281,
+      "learning_rate": 9.564356435643565e-07,
+      "loss": 0.0008,
+      "step": 877
+    },
+    {
+      "epoch": 4.30920245398773,
+      "grad_norm": 0.015077208168804646,
+      "learning_rate": 9.495049504950496e-07,
+      "loss": 0.0012,
+      "step": 878
+    },
+    {
+      "epoch": 4.314110429447853,
+      "grad_norm": 0.029076164588332176,
+      "learning_rate": 9.425742574257426e-07,
+      "loss": 0.0015,
+      "step": 879
+    },
+    {
+      "epoch": 4.319018404907975,
+      "grad_norm": 0.0363786481320858,
+      "learning_rate": 9.356435643564356e-07,
+      "loss": 0.001,
+      "step": 880
+    },
+    {
+      "epoch": 4.323926380368098,
+      "grad_norm": 0.21520403027534485,
+      "learning_rate": 9.287128712871287e-07,
+      "loss": 0.0094,
+      "step": 881
+    },
+    {
+      "epoch": 4.328834355828221,
+      "grad_norm": 0.003572958754375577,
+      "learning_rate": 9.217821782178218e-07,
+      "loss": 0.0005,
+      "step": 882
+    },
+    {
+      "epoch": 4.333742331288343,
+      "grad_norm": 0.01643703132867813,
+      "learning_rate": 9.148514851485148e-07,
+      "loss": 0.0009,
+      "step": 883
+    },
+    {
+      "epoch": 4.338650306748466,
+      "grad_norm": 0.13475348055362701,
+      "learning_rate": 9.079207920792079e-07,
+      "loss": 0.0077,
+      "step": 884
+    },
+    {
+      "epoch": 4.343558282208589,
+      "grad_norm": 0.10863371193408966,
+      "learning_rate": 9.00990099009901e-07,
+      "loss": 0.0294,
+      "step": 885
+    },
+    {
+      "epoch": 4.348466257668711,
+      "grad_norm": 0.02765970304608345,
+      "learning_rate": 8.94059405940594e-07,
+      "loss": 0.0017,
+      "step": 886
+    },
+    {
+      "epoch": 4.353374233128834,
+      "grad_norm": 0.011608476750552654,
+      "learning_rate": 8.871287128712871e-07,
+      "loss": 0.001,
+      "step": 887
+    },
+    {
+      "epoch": 4.358282208588957,
+      "grad_norm": 0.005024611949920654,
+      "learning_rate": 8.801980198019802e-07,
+      "loss": 0.0006,
+      "step": 888
+    },
+    {
+      "epoch": 4.363190184049079,
+      "grad_norm": 0.007748506963253021,
+      "learning_rate": 8.732673267326733e-07,
+      "loss": 0.0006,
+      "step": 889
+    },
+    {
+      "epoch": 4.368098159509202,
+      "grad_norm": 0.23455409705638885,
+      "learning_rate": 8.663366336633663e-07,
+      "loss": 0.0057,
+      "step": 890
+    },
+    {
+      "epoch": 4.373006134969325,
+      "grad_norm": 0.03380454331636429,
+      "learning_rate": 8.594059405940594e-07,
+      "loss": 0.0011,
+      "step": 891
+    },
+    {
+      "epoch": 4.3779141104294474,
+      "grad_norm": 0.03481479734182358,
+      "learning_rate": 8.524752475247525e-07,
+      "loss": 0.0016,
+      "step": 892
+    },
+    {
+      "epoch": 4.38282208588957,
+      "grad_norm": 0.022679351270198822,
+      "learning_rate": 8.455445544554455e-07,
+      "loss": 0.0022,
+      "step": 893
+    },
+    {
+      "epoch": 4.387730061349693,
+      "grad_norm": 0.093803271651268,
+      "learning_rate": 8.386138613861386e-07,
+      "loss": 0.0022,
+      "step": 894
+    },
+    {
+      "epoch": 4.392638036809816,
+      "grad_norm": 0.05329536274075508,
+      "learning_rate": 8.316831683168317e-07,
+      "loss": 0.0018,
+      "step": 895
+    },
+    {
+      "epoch": 4.397546012269939,
+      "grad_norm": 0.05470538139343262,
+      "learning_rate": 8.247524752475247e-07,
+      "loss": 0.0014,
+      "step": 896
+    },
+    {
+      "epoch": 4.402453987730062,
+      "grad_norm": 0.02288208343088627,
+      "learning_rate": 8.178217821782179e-07,
+      "loss": 0.0014,
+      "step": 897
+    },
+    {
+      "epoch": 4.407361963190184,
+      "grad_norm": 0.04405367746949196,
+      "learning_rate": 8.108910891089109e-07,
+      "loss": 0.001,
+      "step": 898
+    },
+    {
+      "epoch": 4.412269938650307,
+      "grad_norm": 0.030512019991874695,
+      "learning_rate": 8.039603960396039e-07,
+      "loss": 0.0013,
+      "step": 899
+    },
+    {
+      "epoch": 4.41717791411043,
+      "grad_norm": 0.12844492495059967,
+      "learning_rate": 7.97029702970297e-07,
+      "loss": 0.0043,
+      "step": 900
+    },
+    {
+      "epoch": 4.422085889570552,
+      "grad_norm": 0.02055547758936882,
+      "learning_rate": 7.900990099009901e-07,
+      "loss": 0.001,
+      "step": 901
+    },
+    {
+      "epoch": 4.426993865030675,
+      "grad_norm": 0.04747156798839569,
+      "learning_rate": 7.831683168316831e-07,
+      "loss": 0.0061,
+      "step": 902
+    },
+    {
+      "epoch": 4.431901840490798,
+      "grad_norm": 0.010140195488929749,
+      "learning_rate": 7.762376237623762e-07,
+      "loss": 0.001,
+      "step": 903
+    },
+    {
+      "epoch": 4.43680981595092,
+      "grad_norm": 0.009971629828214645,
+      "learning_rate": 7.693069306930693e-07,
+      "loss": 0.0011,
+      "step": 904
+    },
+    {
+      "epoch": 4.441717791411043,
+      "grad_norm": 0.006146845407783985,
+      "learning_rate": 7.623762376237624e-07,
+      "loss": 0.0004,
+      "step": 905
+    },
+    {
+      "epoch": 4.446625766871166,
+      "grad_norm": 0.027412964031100273,
+      "learning_rate": 7.554455445544554e-07,
+      "loss": 0.0024,
+      "step": 906
+    },
+    {
+      "epoch": 4.451533742331288,
+      "grad_norm": 0.021934248507022858,
+      "learning_rate": 7.485148514851485e-07,
+      "loss": 0.0012,
+      "step": 907
+    },
+    {
+      "epoch": 4.456441717791411,
+      "grad_norm": 0.3181805908679962,
+      "learning_rate": 7.415841584158416e-07,
+      "loss": 0.0261,
+      "step": 908
+    },
+    {
+      "epoch": 4.461349693251534,
+      "grad_norm": 0.010769632644951344,
+      "learning_rate": 7.346534653465346e-07,
+      "loss": 0.0008,
+      "step": 909
+    },
+    {
+      "epoch": 4.466257668711656,
+      "grad_norm": 0.15605410933494568,
+      "learning_rate": 7.277227722772277e-07,
+      "loss": 0.006,
+      "step": 910
+    },
+    {
+      "epoch": 4.471165644171779,
+      "grad_norm": 0.1743585467338562,
+      "learning_rate": 7.207920792079208e-07,
+      "loss": 0.0075,
+      "step": 911
+    },
+    {
+      "epoch": 4.476073619631902,
+      "grad_norm": 0.011531657539308071,
+      "learning_rate": 7.138613861386138e-07,
+      "loss": 0.0008,
+      "step": 912
+    },
+    {
+      "epoch": 4.480981595092024,
+      "grad_norm": 0.15488475561141968,
+      "learning_rate": 7.06930693069307e-07,
+      "loss": 0.0054,
+      "step": 913
+    },
+    {
+      "epoch": 4.485889570552147,
+      "grad_norm": 0.12285412847995758,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 0.0067,
+      "step": 914
+    },
+    {
+      "epoch": 4.49079754601227,
+      "grad_norm": 0.02667032927274704,
+      "learning_rate": 6.930693069306931e-07,
+      "loss": 0.001,
+      "step": 915
+    },
+    {
+      "epoch": 4.495705521472392,
+      "grad_norm": 0.027680931612849236,
+      "learning_rate": 6.861386138613861e-07,
+      "loss": 0.0014,
+      "step": 916
+    },
+    {
+      "epoch": 4.500613496932515,
+      "grad_norm": 0.01782669499516487,
+      "learning_rate": 6.792079207920792e-07,
+      "loss": 0.0019,
+      "step": 917
+    },
+    {
+      "epoch": 4.505521472392638,
+      "grad_norm": 0.061316560953855515,
+      "learning_rate": 6.722772277227722e-07,
+      "loss": 0.0026,
+      "step": 918
+    },
+    {
+      "epoch": 4.5104294478527605,
+      "grad_norm": 0.052529476583004,
+      "learning_rate": 6.653465346534653e-07,
+      "loss": 0.0023,
+      "step": 919
+    },
+    {
+      "epoch": 4.515337423312883,
+      "grad_norm": 0.0037185668479651213,
+      "learning_rate": 6.584158415841584e-07,
+      "loss": 0.001,
+      "step": 920
+    },
+    {
+      "epoch": 4.520245398773006,
+      "grad_norm": 0.06986022740602493,
+      "learning_rate": 6.514851485148514e-07,
+      "loss": 0.0022,
+      "step": 921
+    },
+    {
+      "epoch": 4.5251533742331285,
+      "grad_norm": 0.006407030858099461,
+      "learning_rate": 6.445544554455445e-07,
+      "loss": 0.0005,
+      "step": 922
+    },
+    {
+      "epoch": 4.530061349693252,
+      "grad_norm": 0.005136528518050909,
+      "learning_rate": 6.376237623762376e-07,
+      "loss": 0.0007,
+      "step": 923
+    },
+    {
+      "epoch": 4.534969325153375,
+      "grad_norm": 0.063414067029953,
+      "learning_rate": 6.306930693069307e-07,
+      "loss": 0.0031,
+      "step": 924
+    },
+    {
+      "epoch": 4.539877300613497,
+      "grad_norm": 0.029516983777284622,
+      "learning_rate": 6.237623762376237e-07,
+      "loss": 0.0015,
+      "step": 925
+    },
+    {
+      "epoch": 4.54478527607362,
+      "grad_norm": 0.07440595328807831,
+      "learning_rate": 6.168316831683168e-07,
+      "loss": 0.0016,
+      "step": 926
+    },
+    {
+      "epoch": 4.549693251533743,
+      "grad_norm": 0.044622018933296204,
+      "learning_rate": 6.099009900990099e-07,
+      "loss": 0.0023,
+      "step": 927
+    },
+    {
+      "epoch": 4.554601226993865,
+      "grad_norm": 0.04841621220111847,
+      "learning_rate": 6.029702970297029e-07,
+      "loss": 0.0049,
+      "step": 928
+    },
+    {
+      "epoch": 4.559509202453988,
+      "grad_norm": 0.026555247604846954,
+      "learning_rate": 5.960396039603961e-07,
+      "loss": 0.001,
+      "step": 929
+    },
+    {
+      "epoch": 4.564417177914111,
+      "grad_norm": 0.1271572858095169,
+      "learning_rate": 5.891089108910892e-07,
+      "loss": 0.0026,
+      "step": 930
+    },
+    {
+      "epoch": 4.569325153374233,
+      "grad_norm": 0.015365286730229855,
+      "learning_rate": 5.821782178217822e-07,
+      "loss": 0.0017,
+      "step": 931
+    },
+    {
+      "epoch": 4.574233128834356,
+      "grad_norm": 0.024656543508172035,
+      "learning_rate": 5.752475247524753e-07,
+      "loss": 0.0013,
+      "step": 932
+    },
+    {
+      "epoch": 4.579141104294479,
+      "grad_norm": 0.05672885477542877,
+      "learning_rate": 5.683168316831684e-07,
+      "loss": 0.0029,
+      "step": 933
+    },
+    {
+      "epoch": 4.584049079754601,
+      "grad_norm": 0.023147093132138252,
+      "learning_rate": 5.613861386138613e-07,
+      "loss": 0.0026,
+      "step": 934
+    },
+    {
+      "epoch": 4.588957055214724,
+      "grad_norm": 0.13222621381282806,
+      "learning_rate": 5.544554455445544e-07,
+      "loss": 0.0144,
+      "step": 935
+    },
+    {
+      "epoch": 4.593865030674847,
+      "grad_norm": 0.022834930568933487,
+      "learning_rate": 5.475247524752475e-07,
+      "loss": 0.0012,
+      "step": 936
+    },
+    {
+      "epoch": 4.598773006134969,
+      "grad_norm": 0.014577291905879974,
+      "learning_rate": 5.405940594059405e-07,
+      "loss": 0.0009,
+      "step": 937
+    },
+    {
+      "epoch": 4.603680981595092,
+      "grad_norm": 0.026476260274648666,
+      "learning_rate": 5.336633663366336e-07,
+      "loss": 0.0015,
+      "step": 938
+    },
+    {
+      "epoch": 4.608588957055215,
+      "grad_norm": 0.0048033553175628185,
+      "learning_rate": 5.267326732673267e-07,
+      "loss": 0.0001,
+      "step": 939
+    },
+    {
+      "epoch": 4.613496932515337,
+      "grad_norm": 0.05348242446780205,
+      "learning_rate": 5.198019801980198e-07,
+      "loss": 0.0017,
+      "step": 940
+    },
+    {
+      "epoch": 4.61840490797546,
+      "grad_norm": 0.027225324884057045,
+      "learning_rate": 5.128712871287128e-07,
+      "loss": 0.0012,
+      "step": 941
+    },
+    {
+      "epoch": 4.623312883435583,
+      "grad_norm": 0.10507699847221375,
+      "learning_rate": 5.059405940594059e-07,
+      "loss": 0.0071,
+      "step": 942
+    },
+    {
+      "epoch": 4.6282208588957054,
+      "grad_norm": 0.12814861536026,
+      "learning_rate": 4.99009900990099e-07,
+      "loss": 0.0073,
+      "step": 943
+    },
+    {
+      "epoch": 4.633128834355828,
+      "grad_norm": 0.03401525318622589,
+      "learning_rate": 4.920792079207921e-07,
+      "loss": 0.0014,
+      "step": 944
+    },
+    {
+      "epoch": 4.638036809815951,
+      "grad_norm": 0.021818110719323158,
+      "learning_rate": 4.851485148514852e-07,
+      "loss": 0.001,
+      "step": 945
+    },
+    {
+      "epoch": 4.6429447852760735,
+      "grad_norm": 0.13617785274982452,
+      "learning_rate": 4.782178217821783e-07,
+      "loss": 0.0047,
+      "step": 946
+    },
+    {
+      "epoch": 4.647852760736196,
+      "grad_norm": 0.15220874547958374,
+      "learning_rate": 4.712871287128713e-07,
+      "loss": 0.0053,
+      "step": 947
+    },
+    {
+      "epoch": 4.652760736196319,
+      "grad_norm": 0.30194413661956787,
+      "learning_rate": 4.6435643564356435e-07,
+      "loss": 0.0218,
+      "step": 948
+    },
+    {
+      "epoch": 4.6576687116564415,
+      "grad_norm": 0.1869322508573532,
+      "learning_rate": 4.574257425742574e-07,
+      "loss": 0.0622,
+      "step": 949
+    },
+    {
+      "epoch": 4.662576687116564,
+      "grad_norm": 0.024376358836889267,
+      "learning_rate": 4.504950495049505e-07,
+      "loss": 0.001,
+      "step": 950
+    },
+    {
+      "epoch": 4.667484662576687,
+      "grad_norm": 0.04647885262966156,
+      "learning_rate": 4.4356435643564356e-07,
+      "loss": 0.0019,
+      "step": 951
+    },
+    {
+      "epoch": 4.6723926380368095,
+      "grad_norm": 0.09357151389122009,
+      "learning_rate": 4.3663366336633663e-07,
+      "loss": 0.0046,
+      "step": 952
+    },
+    {
+      "epoch": 4.677300613496932,
+      "grad_norm": 0.12137161940336227,
+      "learning_rate": 4.297029702970297e-07,
+      "loss": 0.0024,
+      "step": 953
+    },
+    {
+      "epoch": 4.682208588957055,
+      "grad_norm": 0.017510604113340378,
+      "learning_rate": 4.2277227722772276e-07,
+      "loss": 0.0007,
+      "step": 954
+    },
+    {
+      "epoch": 4.6871165644171775,
+      "grad_norm": 0.011096769012510777,
+      "learning_rate": 4.1584158415841583e-07,
+      "loss": 0.0009,
+      "step": 955
+    },
+    {
+      "epoch": 4.6920245398773,
+      "grad_norm": 0.075267493724823,
+      "learning_rate": 4.0891089108910895e-07,
+      "loss": 0.0019,
+      "step": 956
+    },
+    {
+      "epoch": 4.696932515337423,
+      "grad_norm": 0.0064629483968019485,
+      "learning_rate": 4.0198019801980197e-07,
+      "loss": 0.0017,
+      "step": 957
+    },
+    {
+      "epoch": 4.7018404907975455,
+      "grad_norm": 0.018965771421790123,
+      "learning_rate": 3.9504950495049504e-07,
+      "loss": 0.0021,
+      "step": 958
+    },
+    {
+      "epoch": 4.706748466257669,
+      "grad_norm": 0.06010276451706886,
+      "learning_rate": 3.881188118811881e-07,
+      "loss": 0.0014,
+      "step": 959
+    },
+    {
+      "epoch": 4.711656441717792,
+      "grad_norm": 0.03155827522277832,
+      "learning_rate": 3.811881188118812e-07,
+      "loss": 0.0017,
+      "step": 960
+    },
+    {
+      "epoch": 4.716564417177914,
+      "grad_norm": 0.1321091651916504,
+      "learning_rate": 3.7425742574257424e-07,
+      "loss": 0.0066,
+      "step": 961
+    },
+    {
+      "epoch": 4.721472392638037,
+      "grad_norm": 0.018097804859280586,
+      "learning_rate": 3.673267326732673e-07,
+      "loss": 0.0011,
+      "step": 962
+    },
+    {
+      "epoch": 4.72638036809816,
+      "grad_norm": 0.05625467747449875,
+      "learning_rate": 3.603960396039604e-07,
+      "loss": 0.0021,
+      "step": 963
+    },
+    {
+      "epoch": 4.731288343558282,
+      "grad_norm": 0.010586952790617943,
+      "learning_rate": 3.534653465346535e-07,
+      "loss": 0.0004,
+      "step": 964
+    },
+    {
+      "epoch": 4.736196319018405,
+      "grad_norm": 0.0075930338352918625,
+      "learning_rate": 3.4653465346534657e-07,
+      "loss": 0.001,
+      "step": 965
+    },
+    {
+      "epoch": 4.741104294478528,
+      "grad_norm": 0.011664043180644512,
+      "learning_rate": 3.396039603960396e-07,
+      "loss": 0.0014,
+      "step": 966
+    },
+    {
+      "epoch": 4.74601226993865,
+      "grad_norm": 0.04409307986497879,
+      "learning_rate": 3.3267326732673266e-07,
+      "loss": 0.0017,
+      "step": 967
+    },
+    {
+      "epoch": 4.750920245398773,
+      "grad_norm": 0.01488639134913683,
+      "learning_rate": 3.257425742574257e-07,
+      "loss": 0.0012,
+      "step": 968
+    },
+    {
+      "epoch": 4.755828220858896,
+      "grad_norm": 0.02663021720945835,
+      "learning_rate": 3.188118811881188e-07,
+      "loss": 0.0014,
+      "step": 969
+    },
+    {
+      "epoch": 4.7607361963190185,
+      "grad_norm": 0.019124912098050117,
+      "learning_rate": 3.1188118811881186e-07,
+      "loss": 0.0013,
+      "step": 970
+    },
+    {
+      "epoch": 4.765644171779141,
+      "grad_norm": 0.023948566988110542,
+      "learning_rate": 3.0495049504950493e-07,
+      "loss": 0.0027,
+      "step": 971
+    },
+    {
+      "epoch": 4.770552147239264,
+      "grad_norm": 0.04264827072620392,
+      "learning_rate": 2.9801980198019805e-07,
+      "loss": 0.0015,
+      "step": 972
+    },
+    {
+      "epoch": 4.7754601226993865,
+      "grad_norm": 0.11865667253732681,
+      "learning_rate": 2.910891089108911e-07,
+      "loss": 0.0037,
+      "step": 973
+    },
+    {
+      "epoch": 4.780368098159509,
+      "grad_norm": 0.0415462963283062,
+      "learning_rate": 2.841584158415842e-07,
+      "loss": 0.0027,
+      "step": 974
+    },
+    {
+      "epoch": 4.785276073619632,
+      "grad_norm": 0.16526491940021515,
+      "learning_rate": 2.772277227722772e-07,
+      "loss": 0.0149,
+      "step": 975
+    },
+    {
+      "epoch": 4.7901840490797545,
+      "grad_norm": 0.017350684851408005,
+      "learning_rate": 2.7029702970297027e-07,
+      "loss": 0.0008,
+      "step": 976
+    },
+    {
+      "epoch": 4.795092024539877,
+      "grad_norm": 0.020096784457564354,
+      "learning_rate": 2.6336633663366334e-07,
+      "loss": 0.0018,
+      "step": 977
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.019672604277729988,
+      "learning_rate": 2.564356435643564e-07,
+      "loss": 0.0009,
+      "step": 978
+    },
+    {
+      "epoch": 4.8049079754601225,
+      "grad_norm": 0.050094954669475555,
+      "learning_rate": 2.495049504950495e-07,
+      "loss": 0.0015,
+      "step": 979
+    },
+    {
+      "epoch": 4.809815950920245,
+      "grad_norm": 0.07061317563056946,
+      "learning_rate": 2.425742574257426e-07,
+      "loss": 0.0039,
+      "step": 980
+    },
+    {
+      "epoch": 4.814723926380368,
+      "grad_norm": 0.023730693385004997,
+      "learning_rate": 2.3564356435643564e-07,
+      "loss": 0.0012,
+      "step": 981
+    },
+    {
+      "epoch": 4.8196319018404905,
+      "grad_norm": 0.02949446439743042,
+      "learning_rate": 2.287128712871287e-07,
+      "loss": 0.0013,
+      "step": 982
+    },
+    {
+      "epoch": 4.824539877300613,
+      "grad_norm": 0.04460925608873367,
+      "learning_rate": 2.2178217821782178e-07,
+      "loss": 0.002,
+      "step": 983
+    },
+    {
+      "epoch": 4.829447852760736,
+      "grad_norm": 0.032420773059129715,
+      "learning_rate": 2.1485148514851485e-07,
+      "loss": 0.0011,
+      "step": 984
+    },
+    {
+      "epoch": 4.8343558282208585,
+      "grad_norm": 0.16345536708831787,
+      "learning_rate": 2.0792079207920792e-07,
+      "loss": 0.0052,
+      "step": 985
+    },
+    {
+      "epoch": 4.839263803680982,
+      "grad_norm": 0.016612550243735313,
+      "learning_rate": 2.0099009900990098e-07,
+      "loss": 0.0014,
+      "step": 986
+    },
+    {
+      "epoch": 4.844171779141105,
+      "grad_norm": 0.022284861654043198,
+      "learning_rate": 1.9405940594059405e-07,
+      "loss": 0.0015,
+      "step": 987
+    },
+    {
+      "epoch": 4.849079754601227,
+      "grad_norm": 0.018358217552304268,
+      "learning_rate": 1.8712871287128712e-07,
+      "loss": 0.001,
+      "step": 988
+    },
+    {
+      "epoch": 4.85398773006135,
+      "grad_norm": 0.012662719935178757,
+      "learning_rate": 1.801980198019802e-07,
+      "loss": 0.0009,
+      "step": 989
+    },
+    {
+      "epoch": 4.858895705521473,
+      "grad_norm": 0.020140303298830986,
+      "learning_rate": 1.7326732673267329e-07,
+      "loss": 0.0014,
+      "step": 990
+    },
+    {
+      "epoch": 4.863803680981595,
+      "grad_norm": 0.006523944437503815,
+      "learning_rate": 1.6633663366336633e-07,
+      "loss": 0.0004,
+      "step": 991
+    },
+    {
+      "epoch": 4.868711656441718,
+      "grad_norm": 0.02763935551047325,
+      "learning_rate": 1.594059405940594e-07,
+      "loss": 0.0014,
+      "step": 992
+    },
+    {
+      "epoch": 4.873619631901841,
+      "grad_norm": 0.05238136649131775,
+      "learning_rate": 1.5247524752475246e-07,
+      "loss": 0.0015,
+      "step": 993
+    },
+    {
+      "epoch": 4.8785276073619634,
+      "grad_norm": 0.026852233335375786,
+      "learning_rate": 1.4554455445544556e-07,
+      "loss": 0.0011,
+      "step": 994
+    },
+    {
+      "epoch": 4.883435582822086,
+      "grad_norm": 0.01658753491938114,
+      "learning_rate": 1.386138613861386e-07,
+      "loss": 0.0011,
+      "step": 995
+    },
+    {
+      "epoch": 4.888343558282209,
+      "grad_norm": 0.025397639721632004,
+      "learning_rate": 1.3168316831683167e-07,
+      "loss": 0.0018,
+      "step": 996
+    },
+    {
+      "epoch": 4.8932515337423315,
+      "grad_norm": 0.05166466534137726,
+      "learning_rate": 1.2475247524752474e-07,
+      "loss": 0.0014,
+      "step": 997
+    },
+    {
+      "epoch": 4.898159509202454,
+      "grad_norm": 0.06475819647312164,
+      "learning_rate": 1.1782178217821782e-07,
+      "loss": 0.0017,
+      "step": 998
+    },
+    {
+      "epoch": 4.903067484662577,
+      "grad_norm": 0.03307437151670456,
+      "learning_rate": 1.1089108910891089e-07,
+      "loss": 0.0013,
+      "step": 999
+    },
+    {
+      "epoch": 4.9079754601226995,
+      "grad_norm": 0.014889244921505451,
+      "learning_rate": 1.0396039603960396e-07,
+      "loss": 0.0012,
+      "step": 1000
+    },
+    {
+      "epoch": 4.912883435582822,
+      "grad_norm": 0.02157243713736534,
+      "learning_rate": 9.702970297029703e-08,
+      "loss": 0.0013,
+      "step": 1001
+    },
+    {
+      "epoch": 4.917791411042945,
+      "grad_norm": 0.011875314638018608,
+      "learning_rate": 9.00990099009901e-08,
+      "loss": 0.0009,
+      "step": 1002
+    },
+    {
+      "epoch": 4.9226993865030675,
+      "grad_norm": 0.01703159138560295,
+      "learning_rate": 8.316831683168316e-08,
+      "loss": 0.0015,
+      "step": 1003
+    },
+    {
+      "epoch": 4.92760736196319,
+      "grad_norm": 0.042403411120176315,
+      "learning_rate": 7.623762376237623e-08,
+      "loss": 0.0024,
+      "step": 1004
+    },
+    {
+      "epoch": 4.932515337423313,
+      "grad_norm": 0.007990752346813679,
+      "learning_rate": 6.93069306930693e-08,
+      "loss": 0.0011,
+      "step": 1005
+    },
+    {
+      "epoch": 4.9374233128834355,
+      "grad_norm": 0.024830589070916176,
+      "learning_rate": 6.237623762376237e-08,
+      "loss": 0.0013,
+      "step": 1006
+    },
+    {
+      "epoch": 4.942331288343558,
+      "grad_norm": 0.023844977840781212,
+      "learning_rate": 5.5445544554455445e-08,
+      "loss": 0.0011,
+      "step": 1007
+    },
+    {
+      "epoch": 4.947239263803681,
+      "grad_norm": 0.0047842771746218204,
+      "learning_rate": 4.8514851485148513e-08,
+      "loss": 0.0002,
+      "step": 1008
+    },
+    {
+      "epoch": 4.9521472392638035,
+      "grad_norm": 0.20831891894340515,
+      "learning_rate": 4.158415841584158e-08,
+      "loss": 0.0057,
+      "step": 1009
+    },
+    {
+      "epoch": 4.957055214723926,
+      "grad_norm": 0.026182973757386208,
+      "learning_rate": 3.465346534653465e-08,
+      "loss": 0.0017,
+      "step": 1010
+    },
+    {
+      "epoch": 4.961963190184049,
+      "grad_norm": 0.03703535720705986,
+      "learning_rate": 2.7722772277227722e-08,
+      "loss": 0.002,
+      "step": 1011
+    },
+    {
+      "epoch": 4.9668711656441715,
+      "grad_norm": 0.09641406685113907,
+      "learning_rate": 2.079207920792079e-08,
+      "loss": 0.0031,
+      "step": 1012
+    },
+    {
+      "epoch": 4.971779141104294,
+      "grad_norm": 0.007641744799911976,
+      "learning_rate": 1.3861386138613861e-08,
+      "loss": 0.0006,
+      "step": 1013
+    },
+    {
+      "epoch": 4.976687116564417,
+      "grad_norm": 0.15017706155776978,
+      "learning_rate": 6.930693069306931e-09,
+      "loss": 0.0045,
+      "step": 1014
+    },
+    {
+      "epoch": 4.9815950920245395,
+      "grad_norm": 0.014618399553000927,
+      "learning_rate": 0.0,
+      "loss": 0.001,
+      "step": 1015
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1015,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2124067535414374e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}